postgresql/src/interfaces/libpq/fe-secure.c

1488 lines
34 KiB
C
Raw Normal View History

UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
/*-------------------------------------------------------------------------
*
* fe-secure.c
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
* functions related to setting up a secure connection to the backend.
* Secure connections are expected to provide confidentiality,
* message integrity and endpoint authentication.
*
*
2010-01-02 17:58:17 +01:00
* Portions Copyright (c) 1996-2010, PostgreSQL Global Development Group
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
* Portions Copyright (c) 1994, Regents of the University of California
*
*
* IDENTIFICATION
2010-01-02 17:58:17 +01:00
* $PostgreSQL: pgsql/src/interfaces/libpq/fe-secure.c,v 1.131 2010/01/02 16:58:12 momjian Exp $
2002-09-04 22:31:48 +02:00
*
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
* NOTES
*
* We don't provide informational callbacks here (like
* info_cb() in be-secure.c), since there's mechanism to
* display that information to the client.
*
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
*-------------------------------------------------------------------------
*/
#include "postgres_fe.h"
#include <signal.h>
#include <fcntl.h>
#include <ctype.h>
#include "libpq-fe.h"
#include "fe-auth.h"
#include "pqsignal.h"
#include "libpq-int.h"
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
#ifdef WIN32
#include "win32.h"
#else
#include <sys/socket.h>
#include <unistd.h>
#include <netdb.h>
#include <netinet/in.h>
#ifdef HAVE_NETINET_TCP_H
#include <netinet/tcp.h>
#endif
#include <arpa/inet.h>
#endif
#include <sys/stat.h>
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
#ifdef ENABLE_THREAD_SAFETY
#ifdef WIN32
#include "pthread-win32.h"
#else
#include <pthread.h>
#endif
#endif
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
#ifdef USE_SSL
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
#include <openssl/ssl.h>
#include <openssl/bio.h>
#if (SSLEAY_VERSION_NUMBER >= 0x00907000L)
2007-11-15 22:14:46 +01:00
#include <openssl/conf.h>
#endif
#ifdef USE_SSL_ENGINE
#include <openssl/engine.h>
#endif
#ifndef WIN32
2006-04-27 16:02:36 +02:00
#define USER_CERT_FILE ".postgresql/postgresql.crt"
#define USER_KEY_FILE ".postgresql/postgresql.key"
2006-04-27 16:02:36 +02:00
#define ROOT_CERT_FILE ".postgresql/root.crt"
#define ROOT_CRL_FILE ".postgresql/root.crl"
#else
/* On Windows, the "home" directory is already PostgreSQL-specific */
2006-04-27 16:02:36 +02:00
#define USER_CERT_FILE "postgresql.crt"
#define USER_KEY_FILE "postgresql.key"
2006-04-27 16:02:36 +02:00
#define ROOT_CERT_FILE "root.crt"
#define ROOT_CRL_FILE "root.crl"
#endif
2003-08-04 02:43:34 +02:00
#ifndef HAVE_ERR_SET_MARK
/* These don't exist in OpenSSL before 0.9.8 */
#define ERR_set_mark() ((void) 0)
#define ERR_pop_to_mark() ((void) 0)
#endif
static bool verify_peer_name_matches_certificate(PGconn *);
static int verify_cb(int ok, X509_STORE_CTX *ctx);
2002-09-04 22:31:48 +02:00
static int client_cert_cb(SSL *, X509 **, EVP_PKEY **);
static int init_ssl_system(PGconn *conn);
static void destroy_ssl_system(void);
2002-09-04 22:31:48 +02:00
static int initialize_SSL(PGconn *);
static void destroySSL(void);
static PostgresPollingStatusType open_client_SSL(PGconn *);
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
static void close_SSL(PGconn *);
static char *SSLerrmessage(void);
static void SSLerrfree(char *buf);
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
static bool pq_init_ssl_lib = true;
static bool pq_init_crypto_lib = true;
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
static SSL_CTX *SSL_context = NULL;
#ifdef ENABLE_THREAD_SAFETY
static long ssl_open_connections = 0;
#ifndef WIN32
static pthread_mutex_t ssl_config_mutex = PTHREAD_MUTEX_INITIALIZER;
#else
static pthread_mutex_t ssl_config_mutex = NULL;
static long win32_ssl_create_mutex = 0;
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
#endif
#endif /* ENABLE_THREAD_SAFETY */
#endif /* SSL */
/*
* Macros to handle disabling and then restoring the state of SIGPIPE handling.
* On Windows, these are all no-ops since there's no SIGPIPEs.
*/
#ifndef WIN32
#define SIGPIPE_MASKED(conn) ((conn)->sigpipe_so || (conn)->sigpipe_flag)
#ifdef ENABLE_THREAD_SAFETY
struct sigpipe_info
{
sigset_t oldsigmask;
bool sigpipe_pending;
bool got_epipe;
};
#define DECLARE_SIGPIPE_INFO(spinfo) struct sigpipe_info spinfo
#define DISABLE_SIGPIPE(conn, spinfo, failaction) \
do { \
(spinfo).got_epipe = false; \
if (!SIGPIPE_MASKED(conn)) \
{ \
if (pq_block_sigpipe(&(spinfo).oldsigmask, \
&(spinfo).sigpipe_pending) < 0) \
failaction; \
} \
} while (0)
#define REMEMBER_EPIPE(spinfo, cond) \
do { \
if (cond) \
(spinfo).got_epipe = true; \
} while (0)
#define RESTORE_SIGPIPE(conn, spinfo) \
do { \
if (!SIGPIPE_MASKED(conn)) \
pq_reset_sigpipe(&(spinfo).oldsigmask, (spinfo).sigpipe_pending, \
(spinfo).got_epipe); \
} while (0)
#else /* !ENABLE_THREAD_SAFETY */
#define DECLARE_SIGPIPE_INFO(spinfo) pqsigfunc spinfo = NULL
#define DISABLE_SIGPIPE(conn, spinfo, failaction) \
do { \
if (!SIGPIPE_MASKED(conn)) \
spinfo = pqsignal(SIGPIPE, SIG_IGN); \
} while (0)
#define REMEMBER_EPIPE(spinfo, cond)
#define RESTORE_SIGPIPE(conn, spinfo) \
do { \
if (!SIGPIPE_MASKED(conn)) \
pqsignal(SIGPIPE, spinfo); \
} while (0)
#endif /* ENABLE_THREAD_SAFETY */
#else /* WIN32 */
#define DECLARE_SIGPIPE_INFO(spinfo)
#define DISABLE_SIGPIPE(conn, spinfo, failaction)
#define REMEMBER_EPIPE(spinfo, cond)
#define RESTORE_SIGPIPE(conn, spinfo)
#endif /* WIN32 */
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
/* ------------------------------------------------------------ */
2002-09-04 22:31:48 +02:00
/* Procedures common to all secure sessions */
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
/* ------------------------------------------------------------ */
/*
* Exported function to allow application to tell us it's already
* initialized OpenSSL.
*/
void
PQinitSSL(int do_init)
{
PQinitOpenSSL(do_init, do_init);
}
/*
* Exported function to allow application to tell us it's already
* initialized OpenSSL and/or libcrypto.
*/
void
PQinitOpenSSL(int do_ssl, int do_crypto)
{
#ifdef USE_SSL
#ifdef ENABLE_THREAD_SAFETY
/*
* Disallow changing the flags while we have open connections, else we'd
* get completely confused.
*/
if (ssl_open_connections != 0)
return;
#endif
pq_init_ssl_lib = do_ssl;
pq_init_crypto_lib = do_crypto;
#endif
}
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
/*
* Initialize global context
*/
int
2002-09-04 22:31:48 +02:00
pqsecure_initialize(PGconn *conn)
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
{
2002-09-04 22:31:48 +02:00
int r = 0;
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
#ifdef USE_SSL
r = initialize_SSL(conn);
#endif
return r;
}
/*
* Destroy global context
*/
void
2002-09-04 22:31:48 +02:00
pqsecure_destroy(void)
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
{
#ifdef USE_SSL
destroySSL();
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
#endif
}
/*
* Attempt to negotiate secure session.
*/
PostgresPollingStatusType
2002-09-04 22:31:48 +02:00
pqsecure_open_client(PGconn *conn)
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
{
#ifdef USE_SSL
/* First time through? */
if (conn->ssl == NULL)
{
/* We cannot use MSG_NOSIGNAL to block SIGPIPE when using SSL */
conn->sigpipe_flag = false;
if (!(conn->ssl = SSL_new(SSL_context)) ||
!SSL_set_app_data(conn->ssl, conn) ||
!SSL_set_fd(conn->ssl, conn->sock))
{
2004-08-29 07:07:03 +02:00
char *err = SSLerrmessage();
printfPQExpBuffer(&conn->errorMessage,
2005-10-15 04:49:52 +02:00
libpq_gettext("could not establish SSL connection: %s\n"),
err);
SSLerrfree(err);
close_SSL(conn);
return PGRES_POLLING_FAILED;
}
2005-10-15 04:49:52 +02:00
/*
2005-10-15 04:49:52 +02:00
* Initialize errorMessage to empty. This allows open_client_SSL() to
* detect whether client_cert_cb() has stored a message.
*/
resetPQExpBuffer(&conn->errorMessage);
}
/* Begin or continue the actual handshake */
return open_client_SSL(conn);
#else
/* shouldn't get here */
return PGRES_POLLING_FAILED;
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
#endif
}
/*
* Close secure session.
*/
void
2002-09-04 22:31:48 +02:00
pqsecure_close(PGconn *conn)
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
{
#ifdef USE_SSL
if (conn->ssl)
close_SSL(conn);
#endif
}
/*
* Read data from a secure connection.
*/
ssize_t
2002-09-04 22:31:48 +02:00
pqsecure_read(PGconn *conn, void *ptr, size_t len)
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
{
2002-09-04 22:31:48 +02:00
ssize_t n;
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
#ifdef USE_SSL
if (conn->ssl)
{
int err;
DECLARE_SIGPIPE_INFO(spinfo);
/* SSL_read can write to the socket, so we need to disable SIGPIPE */
DISABLE_SIGPIPE(conn, spinfo, return -1);
2003-08-04 02:43:34 +02:00
rloop:
SOCK_ERRNO_SET(0);
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
n = SSL_read(conn->ssl, ptr, len);
err = SSL_get_error(conn->ssl, n);
switch (err)
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
{
2002-09-04 22:31:48 +02:00
case SSL_ERROR_NONE:
break;
case SSL_ERROR_WANT_READ:
n = 0;
break;
case SSL_ERROR_WANT_WRITE:
2003-08-04 02:43:34 +02:00
/*
2005-10-15 04:49:52 +02:00
* Returning 0 here would cause caller to wait for read-ready,
* which is not correct since what SSL wants is wait for
* write-ready. The former could get us stuck in an infinite
* wait, so don't risk it; busy-loop instead.
*/
goto rloop;
2002-09-04 22:31:48 +02:00
case SSL_ERROR_SYSCALL:
2003-08-04 02:43:34 +02:00
{
char sebuf[256];
if (n == -1)
{
REMEMBER_EPIPE(spinfo, SOCK_ERRNO == EPIPE);
2003-08-04 02:43:34 +02:00
printfPQExpBuffer(&conn->errorMessage,
2005-10-15 04:49:52 +02:00
libpq_gettext("SSL SYSCALL error: %s\n"),
SOCK_STRERROR(SOCK_ERRNO, sebuf, sizeof(sebuf)));
}
2003-08-04 02:43:34 +02:00
else
{
2003-08-04 02:43:34 +02:00
printfPQExpBuffer(&conn->errorMessage,
2005-10-15 04:49:52 +02:00
libpq_gettext("SSL SYSCALL error: EOF detected\n"));
SOCK_ERRNO_SET(ECONNRESET);
n = -1;
}
2003-08-04 02:43:34 +02:00
break;
}
2002-09-04 22:31:48 +02:00
case SSL_ERROR_SSL:
{
2004-08-29 07:07:03 +02:00
char *err = SSLerrmessage();
printfPQExpBuffer(&conn->errorMessage,
2005-10-15 04:49:52 +02:00
libpq_gettext("SSL error: %s\n"), err);
SSLerrfree(err);
}
2002-09-04 22:31:48 +02:00
/* fall through */
case SSL_ERROR_ZERO_RETURN:
SOCK_ERRNO_SET(ECONNRESET);
2002-09-04 22:31:48 +02:00
n = -1;
break;
default:
printfPQExpBuffer(&conn->errorMessage,
2005-10-15 04:49:52 +02:00
libpq_gettext("unrecognized SSL error code: %d\n"),
err);
n = -1;
break;
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
}
RESTORE_SIGPIPE(conn, spinfo);
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
}
else
#endif
2002-09-04 22:31:48 +02:00
n = recv(conn->sock, ptr, len, 0);
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
return n;
}
/*
* Write data to a secure connection.
*/
ssize_t
2002-09-04 22:31:48 +02:00
pqsecure_write(PGconn *conn, const void *ptr, size_t len)
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
{
2002-09-04 22:31:48 +02:00
ssize_t n;
DECLARE_SIGPIPE_INFO(spinfo);
2005-10-15 04:49:52 +02:00
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
#ifdef USE_SSL
if (conn->ssl)
{
int err;
DISABLE_SIGPIPE(conn, spinfo, return -1);
SOCK_ERRNO_SET(0);
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
n = SSL_write(conn->ssl, ptr, len);
err = SSL_get_error(conn->ssl, n);
switch (err)
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
{
2002-09-04 22:31:48 +02:00
case SSL_ERROR_NONE:
break;
case SSL_ERROR_WANT_READ:
2003-08-04 02:43:34 +02:00
/*
* Returning 0 here causes caller to wait for write-ready,
2005-10-15 04:49:52 +02:00
* which is not really the right thing, but it's the best we
* can do.
*/
n = 0;
break;
2002-09-04 22:31:48 +02:00
case SSL_ERROR_WANT_WRITE:
n = 0;
break;
2002-09-04 22:31:48 +02:00
case SSL_ERROR_SYSCALL:
2003-08-04 02:43:34 +02:00
{
char sebuf[256];
2003-08-04 02:43:34 +02:00
if (n == -1)
{
REMEMBER_EPIPE(spinfo, SOCK_ERRNO == EPIPE);
2003-08-04 02:43:34 +02:00
printfPQExpBuffer(&conn->errorMessage,
2005-10-15 04:49:52 +02:00
libpq_gettext("SSL SYSCALL error: %s\n"),
SOCK_STRERROR(SOCK_ERRNO, sebuf, sizeof(sebuf)));
}
2003-08-04 02:43:34 +02:00
else
{
2003-08-04 02:43:34 +02:00
printfPQExpBuffer(&conn->errorMessage,
2005-10-15 04:49:52 +02:00
libpq_gettext("SSL SYSCALL error: EOF detected\n"));
SOCK_ERRNO_SET(ECONNRESET);
n = -1;
}
2003-08-04 02:43:34 +02:00
break;
}
2002-09-04 22:31:48 +02:00
case SSL_ERROR_SSL:
{
2004-08-29 07:07:03 +02:00
char *err = SSLerrmessage();
printfPQExpBuffer(&conn->errorMessage,
2005-10-15 04:49:52 +02:00
libpq_gettext("SSL error: %s\n"), err);
SSLerrfree(err);
}
2002-09-04 22:31:48 +02:00
/* fall through */
case SSL_ERROR_ZERO_RETURN:
SOCK_ERRNO_SET(ECONNRESET);
2002-09-04 22:31:48 +02:00
n = -1;
break;
default:
printfPQExpBuffer(&conn->errorMessage,
2005-10-15 04:49:52 +02:00
libpq_gettext("unrecognized SSL error code: %d\n"),
err);
n = -1;
break;
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
}
}
else
#endif
{
int flags = 0;
#ifdef MSG_NOSIGNAL
if (conn->sigpipe_flag)
flags |= MSG_NOSIGNAL;
retry_masked:
#endif /* MSG_NOSIGNAL */
DISABLE_SIGPIPE(conn, spinfo, return -1);
n = send(conn->sock, ptr, len, flags);
if (n < 0)
{
/*
* If we see an EINVAL, it may be because MSG_NOSIGNAL isn't
* available on this machine. So, clear sigpipe_flag so we don't
* try the flag again, and retry the send().
*/
#ifdef MSG_NOSIGNAL
if (flags != 0 && SOCK_ERRNO == EINVAL)
{
conn->sigpipe_flag = false;
flags = 0;
goto retry_masked;
}
#endif /* MSG_NOSIGNAL */
REMEMBER_EPIPE(spinfo, SOCK_ERRNO == EPIPE);
}
}
2005-10-15 04:49:52 +02:00
RESTORE_SIGPIPE(conn, spinfo);
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
return n;
}
/* ------------------------------------------------------------ */
2002-09-04 22:31:48 +02:00
/* SSL specific code */
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
/* ------------------------------------------------------------ */
#ifdef USE_SSL
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
/*
* Certificate verification callback
*
* This callback allows us to log intermediate problems during
* verification, but there doesn't seem to be a clean way to get
* our PGconn * structure. So we can't log anything!
*
* This callback also allows us to override the default acceptance
* criteria (e.g., accepting self-signed or expired certs), but
* for now we accept the default checks.
*/
static int
2002-09-04 22:31:48 +02:00
verify_cb(int ok, X509_STORE_CTX *ctx)
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
{
return ok;
}
/*
* Check if a wildcard certificate matches the server hostname.
*
* The rule for this is:
* 1. We only match the '*' character as wildcard
* 2. We match only wildcards at the start of the string
* 3. The '*' character does *not* match '.', meaning that we match only
* a single pathname component.
* 4. We don't support more than one '*' in a single pattern.
*
* This is roughly in line with RFC2818, but contrary to what most browsers
* appear to be implementing (point 3 being the difference)
*
* Matching is always cone case-insensitive, since DNS is case insensitive.
*/
static int
wildcard_certificate_match(const char *pattern, const char *string)
{
int lenpat = strlen(pattern);
int lenstr = strlen(string);
/* If we don't start with a wildcard, it's not a match (rule 1 & 2) */
if (lenpat < 3 ||
pattern[0] != '*' ||
pattern[1] != '.')
return 0;
if (lenpat > lenstr)
/* If pattern is longer than the string, we can never match */
return 0;
if (pg_strcasecmp(pattern + 1, string + lenstr - lenpat + 1) != 0)
/*
* If string does not end in pattern (minus the wildcard), we don't
* match
*/
return 0;
if (strchr(string, '.') < string + lenstr - lenpat)
/*
* If there is a dot left of where the pattern started to match, we
* don't match (rule 3)
*/
return 0;
/* String ended with pattern, and didn't have a dot before, so we match */
return 1;
}
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
/*
* Verify that common name resolves to peer.
*/
static bool
verify_peer_name_matches_certificate(PGconn *conn)
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
{
/*
* If told not to verify the peer name, don't do it. Return 0 indicating
* that the verification was successful.
*/
if (strcmp(conn->sslmode, "verify-full") != 0)
return true;
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
if (conn->pghostaddr)
{
printfPQExpBuffer(&conn->errorMessage,
2009-01-19 09:59:13 +01:00
libpq_gettext("verified SSL connections are only supported when connecting to a host name"));
return false;
}
else
{
/*
* Connect by hostname.
*
* XXX: Should support alternate names here
*/
if (pg_strcasecmp(conn->peer_cn, conn->pghost) == 0)
/* Exact name match */
return true;
else if (wildcard_certificate_match(conn->peer_cn, conn->pghost))
/* Matched wildcard certificate */
return true;
else
{
printfPQExpBuffer(&conn->errorMessage,
2009-01-19 09:59:13 +01:00
libpq_gettext("server common name \"%s\" does not match host name \"%s\""),
conn->peer_cn, conn->pghost);
return false;
}
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
}
}
/*
* Callback used by SSL to load client cert and key.
* This callback is only called when the server wants a
* client cert.
*
* Since BIO functions can set OpenSSL error codes, we must
* reset the OpenSSL error stack on *every* exit from this
* function once we've started using BIO.
*
* Must return 1 on success, 0 on no data or error.
*/
static int
2002-09-04 22:31:48 +02:00
client_cert_cb(SSL *ssl, X509 **x509, EVP_PKEY **pkey)
{
char homedir[MAXPGPATH];
struct stat buf;
2005-10-15 04:49:52 +02:00
#ifndef WIN32
struct stat buf2;
2007-11-15 22:14:46 +01:00
FILE *fp;
#endif
char fnbuf[MAXPGPATH];
2007-11-15 22:14:46 +01:00
BIO *bio;
PGconn *conn = (PGconn *) SSL_get_app_data(ssl);
2003-08-04 02:43:34 +02:00
char sebuf[256];
/*
* If conn->sslcert or conn->sslkey is not set, we don't need the home
* directory to find the required files.
*/
if (!conn->sslcert || !conn->sslkey)
{
if (!pqGetHomeDirectory(homedir, sizeof(homedir)))
{
printfPQExpBuffer(&conn->errorMessage,
2009-01-19 09:59:13 +01:00
libpq_gettext("could not get home directory to locate client certificate files"));
return 0;
}
}
/* read the user certificate */
if (conn->sslcert)
strncpy(fnbuf, conn->sslcert, sizeof(fnbuf));
else
snprintf(fnbuf, sizeof(fnbuf), "%s/%s", homedir, USER_CERT_FILE);
2007-11-15 22:14:46 +01:00
/*
* OpenSSL <= 0.9.8 lacks error stack handling, which means it's likely to
* report wrong error messages if access to the cert file fails. Do our
* own check for the readability of the file to catch the majority of such
* problems before OpenSSL gets involved.
*/
#ifndef HAVE_ERR_SET_MARK
{
2007-11-15 22:14:46 +01:00
FILE *fp2;
if ((fp2 = fopen(fnbuf, "r")) == NULL)
{
printfPQExpBuffer(&conn->errorMessage,
2007-11-15 22:14:46 +01:00
libpq_gettext("could not open certificate file \"%s\": %s\n"),
fnbuf, pqStrerror(errno, sebuf, sizeof(sebuf)));
return 0;
}
fclose(fp2);
}
#endif
/* save OpenSSL error stack */
ERR_set_mark();
2007-11-15 22:14:46 +01:00
if ((bio = BIO_new_file(fnbuf, "r")) == NULL)
{
2002-09-04 22:31:48 +02:00
printfPQExpBuffer(&conn->errorMessage,
2005-10-15 04:49:52 +02:00
libpq_gettext("could not open certificate file \"%s\": %s\n"),
fnbuf, pqStrerror(errno, sebuf, sizeof(sebuf)));
ERR_pop_to_mark();
return 0;
}
if (PEM_read_bio_X509(bio, x509, NULL, NULL) == NULL)
{
2004-08-29 07:07:03 +02:00
char *err = SSLerrmessage();
2002-09-04 22:31:48 +02:00
printfPQExpBuffer(&conn->errorMessage,
2005-10-15 04:49:52 +02:00
libpq_gettext("could not read certificate file \"%s\": %s\n"),
fnbuf, err);
SSLerrfree(err);
BIO_free(bio);
ERR_pop_to_mark();
return 0;
}
BIO_free(bio);
/*
* Read the SSL key. If a key is specified, treat it as an engine:key
* combination if there is colon present - we don't support files with
* colon in the name. The exception is if the second character is a colon,
* in which case it can be a Windows filename with drive specification.
*/
if (conn->sslkey && strlen(conn->sslkey) > 0)
{
#ifdef USE_SSL_ENGINE
if (strchr(conn->sslkey, ':')
#ifdef WIN32
&& conn->sslkey[1] != ':'
#endif
)
{
/* Colon, but not in second character, treat as engine:key */
char *engine_str = strdup(conn->sslkey);
char *engine_colon = strchr(engine_str, ':');
*engine_colon = '\0'; /* engine_str now has engine name */
engine_colon++; /* engine_colon now has key name */
conn->engine = ENGINE_by_id(engine_str);
if (conn->engine == NULL)
{
char *err = SSLerrmessage();
2007-02-16 18:07:00 +01:00
printfPQExpBuffer(&conn->errorMessage,
libpq_gettext("could not load SSL engine \"%s\": %s\n"),
engine_str, err);
SSLerrfree(err);
free(engine_str);
ERR_pop_to_mark();
return 0;
}
if (ENGINE_init(conn->engine) == 0)
{
char *err = SSLerrmessage();
printfPQExpBuffer(&conn->errorMessage,
libpq_gettext("could not initialize SSL engine \"%s\": %s\n"),
engine_str, err);
SSLerrfree(err);
ENGINE_free(conn->engine);
conn->engine = NULL;
free(engine_str);
ERR_pop_to_mark();
return 0;
}
*pkey = ENGINE_load_private_key(conn->engine, engine_colon,
NULL, NULL);
if (*pkey == NULL)
{
char *err = SSLerrmessage();
printfPQExpBuffer(&conn->errorMessage,
libpq_gettext("could not read private SSL key \"%s\" from engine \"%s\": %s\n"),
engine_colon, engine_str, err);
SSLerrfree(err);
ENGINE_finish(conn->engine);
ENGINE_free(conn->engine);
conn->engine = NULL;
free(engine_str);
ERR_pop_to_mark();
return 0;
}
free(engine_str);
fnbuf[0] = '\0'; /* indicate we're not going to load from a
* file */
}
else
#endif /* support for SSL engines */
{
/* PGSSLKEY is not an engine, treat it as a filename */
strncpy(fnbuf, conn->sslkey, sizeof(fnbuf));
}
}
else
{
/* No PGSSLKEY specified, load default file */
snprintf(fnbuf, sizeof(fnbuf), "%s/%s", homedir, USER_KEY_FILE);
}
if (fnbuf[0] != '\0')
{
/* read the user key from file */
if (stat(fnbuf, &buf) != 0)
{
printfPQExpBuffer(&conn->errorMessage,
2007-11-15 22:14:46 +01:00
libpq_gettext("certificate present, but not private key file \"%s\"\n"),
fnbuf);
ERR_pop_to_mark();
return 0;
}
2007-02-16 18:07:00 +01:00
#ifndef WIN32
if (!S_ISREG(buf.st_mode) || buf.st_mode & (S_IRWXG | S_IRWXO))
{
printfPQExpBuffer(&conn->errorMessage,
libpq_gettext("private key file \"%s\" has group or world access; permissions should be u=rw (0600) or less\n"),
2007-11-15 22:14:46 +01:00
fnbuf);
ERR_pop_to_mark();
return 0;
}
2007-02-16 18:07:00 +01:00
#endif
if ((bio = BIO_new_file(fnbuf, "r")) == NULL)
{
printfPQExpBuffer(&conn->errorMessage,
2007-11-15 22:14:46 +01:00
libpq_gettext("could not open private key file \"%s\": %s\n"),
fnbuf, pqStrerror(errno, sebuf, sizeof(sebuf)));
ERR_pop_to_mark();
return 0;
}
2007-02-16 18:07:00 +01:00
#ifndef WIN32
BIO_get_fp(bio, &fp);
if (fstat(fileno(fp), &buf2) == -1 ||
buf.st_dev != buf2.st_dev || buf.st_ino != buf2.st_ino)
{
printfPQExpBuffer(&conn->errorMessage,
2007-11-15 22:14:46 +01:00
libpq_gettext("private key file \"%s\" changed during execution\n"), fnbuf);
ERR_pop_to_mark();
return 0;
}
2007-02-16 18:07:00 +01:00
#endif
if (PEM_read_bio_PrivateKey(bio, pkey, NULL, NULL) == NULL)
{
char *err = SSLerrmessage();
2004-08-29 07:07:03 +02:00
printfPQExpBuffer(&conn->errorMessage,
2007-11-15 22:14:46 +01:00
libpq_gettext("could not read private key file \"%s\": %s\n"),
fnbuf, err);
SSLerrfree(err);
BIO_free(bio);
ERR_pop_to_mark();
return 0;
}
BIO_free(bio);
}
2007-02-16 18:07:00 +01:00
/* verify that the cert and key go together */
if (X509_check_private_key(*x509, *pkey) != 1)
{
2004-08-29 07:07:03 +02:00
char *err = SSLerrmessage();
2002-09-04 22:31:48 +02:00
printfPQExpBuffer(&conn->errorMessage,
2005-10-15 04:49:52 +02:00
libpq_gettext("certificate does not match private key file \"%s\": %s\n"),
fnbuf, err);
SSLerrfree(err);
ERR_pop_to_mark();
return 0;
}
ERR_pop_to_mark();
return 1;
}
#ifdef ENABLE_THREAD_SAFETY
/*
* Callback functions for OpenSSL internal locking
*/
static unsigned long
pq_threadidcallback(void)
{
2005-08-28 18:37:48 +02:00
/*
2006-05-12 01:27:35 +02:00
* This is not standards-compliant. pthread_self() returns pthread_t, and
2005-10-15 04:49:52 +02:00
* shouldn't be cast to unsigned long, but CRYPTO_set_id_callback requires
* it, so we have to do it.
2005-08-28 18:37:48 +02:00
*/
2004-08-29 07:07:03 +02:00
return (unsigned long) pthread_self();
}
static pthread_mutex_t *pq_lockarray;
static void
pq_lockingcallback(int mode, int n, const char *file, int line)
{
2004-08-29 07:07:03 +02:00
if (mode & CRYPTO_LOCK)
{
if (pthread_mutex_lock(&pq_lockarray[n]))
PGTHREAD_ERROR("failed to lock mutex");
}
2004-08-29 07:07:03 +02:00
else
{
if (pthread_mutex_unlock(&pq_lockarray[n]))
PGTHREAD_ERROR("failed to unlock mutex");
}
}
2004-08-29 07:07:03 +02:00
#endif /* ENABLE_THREAD_SAFETY */
/*
* Initialize SSL system. In threadsafe mode, this includes setting
* up libcrypto callback functions to do thread locking.
*
* If the caller has told us (through PQinitOpenSSL) that he's taking care
* of libcrypto, we expect that callbacks are already set, and won't try to
* override it.
*
* The conn parameter is only used to be able to pass back an error
* message - no connection local setup is made.
*/
static int
init_ssl_system(PGconn *conn)
{
#ifdef ENABLE_THREAD_SAFETY
#ifdef WIN32
/* Also see similar code in fe-connect.c, default_threadlock() */
if (ssl_config_mutex == NULL)
2004-08-29 07:07:03 +02:00
{
while (InterlockedExchange(&win32_ssl_create_mutex, 1) == 1)
2004-08-29 07:07:03 +02:00
/* loop, another thread own the lock */ ;
if (ssl_config_mutex == NULL)
{
if (pthread_mutex_init(&ssl_config_mutex, NULL))
return -1;
}
InterlockedExchange(&win32_ssl_create_mutex, 0);
}
#endif
if (pthread_mutex_lock(&ssl_config_mutex))
return -1;
2004-08-29 07:07:03 +02:00
if (pq_init_crypto_lib)
2004-08-29 07:07:03 +02:00
{
/*
* If necessary, set up an array to hold locks for libcrypto.
* libcrypto will tell us how big to make this array.
*/
if (pq_lockarray == NULL)
{
int i;
pq_lockarray = malloc(sizeof(pthread_mutex_t) * CRYPTO_num_locks());
if (!pq_lockarray)
{
pthread_mutex_unlock(&ssl_config_mutex);
return -1;
}
for (i = 0; i < CRYPTO_num_locks(); i++)
{
if (pthread_mutex_init(&pq_lockarray[i], NULL))
{
free(pq_lockarray);
pq_lockarray = NULL;
pthread_mutex_unlock(&ssl_config_mutex);
return -1;
}
}
}
if (ssl_open_connections++ == 0)
{
/* These are only required for threaded libcrypto applications */
CRYPTO_set_id_callback(pq_threadidcallback);
CRYPTO_set_locking_callback(pq_lockingcallback);
}
}
#endif /* ENABLE_THREAD_SAFETY */
if (!SSL_context)
{
if (pq_init_ssl_lib)
2004-08-29 07:07:03 +02:00
{
2007-02-16 18:07:00 +01:00
#if SSLEAY_VERSION_NUMBER >= 0x00907000L
OPENSSL_config(NULL);
2007-11-15 22:14:46 +01:00
#endif
SSL_library_init();
SSL_load_error_strings();
}
SSL_context = SSL_CTX_new(TLSv1_method());
if (!SSL_context)
{
2004-08-29 07:07:03 +02:00
char *err = SSLerrmessage();
printfPQExpBuffer(&conn->errorMessage,
2005-10-15 04:49:52 +02:00
libpq_gettext("could not create SSL context: %s\n"),
err);
SSLerrfree(err);
#ifdef ENABLE_THREAD_SAFETY
pthread_mutex_unlock(&ssl_config_mutex);
#endif
return -1;
}
}
#ifdef ENABLE_THREAD_SAFETY
pthread_mutex_unlock(&ssl_config_mutex);
#endif
return 0;
}
2004-08-29 07:07:03 +02:00
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
/*
* This function is needed because if the libpq library is unloaded
* from the application, the callback functions will no longer exist when
* libcrypto is used by other parts of the system. For this reason,
* we unregister the callback functions when the last libpq
* connection is closed. (The same would apply for OpenSSL callbacks
* if we had any.)
*
* Callbacks are only set when we're compiled in threadsafe mode, so
* we only need to remove them in this case.
*/
static void
destroy_ssl_system(void)
{
#ifdef ENABLE_THREAD_SAFETY
/* Mutex is created in initialize_ssl_system() */
if (pthread_mutex_lock(&ssl_config_mutex))
return;
if (pq_init_crypto_lib && ssl_open_connections > 0)
--ssl_open_connections;
if (pq_init_crypto_lib && ssl_open_connections == 0)
{
/* No connections left, unregister libcrypto callbacks */
CRYPTO_set_locking_callback(NULL);
CRYPTO_set_id_callback(NULL);
/*
* We don't free the lock array. If we get another connection in this
* process, we will just re-use it with the existing mutexes.
*
* This means we leak a little memory on repeated load/unload of the
* library.
*/
}
pthread_mutex_unlock(&ssl_config_mutex);
#endif
return;
}
/*
* Initialize SSL context.
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
*/
static int
2002-09-04 22:31:48 +02:00
initialize_SSL(PGconn *conn)
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
{
struct stat buf;
char homedir[MAXPGPATH];
char fnbuf[MAXPGPATH];
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
2004-08-29 07:07:03 +02:00
if (init_ssl_system(conn))
return -1;
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
/*
* If sslmode is set to one of the verify options, perform certificate
* verification. If set to "verify-full" we will also do further
* verification after the connection has been completed.
*
* If we are going to look for either root certificate or CRL in the home
* directory, we need pqGetHomeDirectory() to succeed. In other cases, we
* don't need to get the home directory explicitly.
*/
if (!conn->sslrootcert || !conn->sslcrl)
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
{
if (!pqGetHomeDirectory(homedir, sizeof(homedir)))
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
{
if (conn->sslmode[0] == 'v') /* "verify-ca" or
* "verify-full" */
{
printfPQExpBuffer(&conn->errorMessage,
2009-01-19 09:59:13 +01:00
libpq_gettext("could not get home directory to locate root certificate file"));
return -1;
}
}
}
else
{
homedir[0] = '\0';
}
2004-08-29 07:07:03 +02:00
if (conn->sslrootcert)
strncpy(fnbuf, conn->sslrootcert, sizeof(fnbuf));
else
snprintf(fnbuf, sizeof(fnbuf), "%s/%s", homedir, ROOT_CERT_FILE);
if (stat(fnbuf, &buf) == 0)
{
X509_STORE *cvstore;
if (SSL_CTX_load_verify_locations(SSL_context, fnbuf, NULL) != 1)
{
char *err = SSLerrmessage();
2006-10-04 02:30:14 +02:00
printfPQExpBuffer(&conn->errorMessage,
libpq_gettext("could not read root certificate file \"%s\": %s\n"),
fnbuf, err);
SSLerrfree(err);
return -1;
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
}
if ((cvstore = SSL_CTX_get_cert_store(SSL_context)) != NULL)
{
if (conn->sslcrl)
strncpy(fnbuf, conn->sslcrl, sizeof(fnbuf));
else
snprintf(fnbuf, sizeof(fnbuf), "%s/%s", homedir, ROOT_CRL_FILE);
/* setting the flags to check against the complete CRL chain */
if (X509_STORE_load_locations(cvstore, fnbuf, NULL) == 1)
/* OpenSSL 0.96 does not support X509_V_FLAG_CRL_CHECK */
#ifdef X509_V_FLAG_CRL_CHECK
X509_STORE_set_flags(cvstore,
X509_V_FLAG_CRL_CHECK | X509_V_FLAG_CRL_CHECK_ALL);
/* if not found, silently ignore; we do not require CRL */
#else
{
char *err = SSLerrmessage();
printfPQExpBuffer(&conn->errorMessage,
libpq_gettext("SSL library does not support CRL certificates (file \"%s\")\n"),
fnbuf);
SSLerrfree(err);
return -1;
}
#endif
}
SSL_CTX_set_verify(SSL_context, SSL_VERIFY_PEER, verify_cb);
}
else
{
/* stat() failed; assume cert file doesn't exist */
if (conn->sslmode[0] == 'v') /* "verify-ca" or "verify-full" */
{
printfPQExpBuffer(&conn->errorMessage,
libpq_gettext("root certificate file \"%s\" does not exist\n"
"Either provide the file or change sslmode to disable server certificate verification.\n"), fnbuf);
return -1;
}
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
}
/* set up mechanism to provide client certificate, if available */
SSL_CTX_set_client_cert_cb(SSL_context, client_cert_cb);
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
return 0;
}
static void
destroySSL(void)
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
{
destroy_ssl_system();
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
}
/*
* Attempt to negotiate SSL connection.
*/
static PostgresPollingStatusType
2002-09-04 22:31:48 +02:00
open_client_SSL(PGconn *conn)
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
{
2002-09-04 22:31:48 +02:00
int r;
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
r = SSL_connect(conn->ssl);
if (r <= 0)
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
{
2005-10-15 04:49:52 +02:00
int err = SSL_get_error(conn->ssl, r);
switch (err)
{
case SSL_ERROR_WANT_READ:
return PGRES_POLLING_READING;
2003-08-04 02:43:34 +02:00
case SSL_ERROR_WANT_WRITE:
return PGRES_POLLING_WRITING;
case SSL_ERROR_SYSCALL:
2003-08-04 02:43:34 +02:00
{
char sebuf[256];
if (r == -1)
printfPQExpBuffer(&conn->errorMessage,
2005-10-15 04:49:52 +02:00
libpq_gettext("SSL SYSCALL error: %s\n"),
SOCK_STRERROR(SOCK_ERRNO, sebuf, sizeof(sebuf)));
2003-08-04 02:43:34 +02:00
else
printfPQExpBuffer(&conn->errorMessage,
2005-10-15 04:49:52 +02:00
libpq_gettext("SSL SYSCALL error: EOF detected\n"));
2003-08-04 02:43:34 +02:00
close_SSL(conn);
return PGRES_POLLING_FAILED;
}
case SSL_ERROR_SSL:
{
/*
* If there are problems with the local certificate files,
* these will be detected by client_cert_cb() which is
* called from SSL_connect(). We want to return that
* error message and not the rather unhelpful error that
2005-10-15 04:49:52 +02:00
* OpenSSL itself returns. So check to see if an error
* message was already stored.
*/
if (conn->errorMessage.len == 0)
{
char *err = SSLerrmessage();
2004-08-29 07:07:03 +02:00
printfPQExpBuffer(&conn->errorMessage,
libpq_gettext("SSL error: %s\n"),
err);
SSLerrfree(err);
}
close_SSL(conn);
return PGRES_POLLING_FAILED;
}
default:
printfPQExpBuffer(&conn->errorMessage,
2005-10-15 04:49:52 +02:00
libpq_gettext("unrecognized SSL error code: %d\n"),
err);
close_SSL(conn);
return PGRES_POLLING_FAILED;
}
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
}
2002-09-04 22:31:48 +02:00
/*
* We already checked the server certificate in initialize_SSL() using
* SSL_CTX_set_verify() if root.crt exists.
2002-09-04 22:31:48 +02:00
*/
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
/* pull out server distinguished and common names */
conn->peer = SSL_get_peer_certificate(conn->ssl);
if (conn->peer == NULL)
{
2004-08-29 07:07:03 +02:00
char *err = SSLerrmessage();
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
printfPQExpBuffer(&conn->errorMessage,
2005-10-15 04:49:52 +02:00
libpq_gettext("certificate could not be obtained: %s\n"),
err);
SSLerrfree(err);
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
close_SSL(conn);
return PGRES_POLLING_FAILED;
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
}
X509_NAME_oneline(X509_get_subject_name(conn->peer),
2002-09-04 22:31:48 +02:00
conn->peer_dn, sizeof(conn->peer_dn));
conn->peer_dn[sizeof(conn->peer_dn) - 1] = '\0';
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
r = X509_NAME_get_text_by_NID(X509_get_subject_name(conn->peer),
2002-09-04 22:31:48 +02:00
NID_commonName, conn->peer_cn, SM_USER);
conn->peer_cn[SM_USER] = '\0'; /* buffer is SM_USER+1 chars! */
if (r == -1)
{
/* Unable to get the CN, set it to blank so it can't be used */
conn->peer_cn[0] = '\0';
}
else
{
/*
* Reject embedded NULLs in certificate common name to prevent attacks like
* CVE-2009-4034.
*/
if (r != strlen(conn->peer_cn))
{
printfPQExpBuffer(&conn->errorMessage,
libpq_gettext("SSL certificate's common name contains embedded null\n"));
close_SSL(conn);
return PGRES_POLLING_FAILED;
}
}
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
if (!verify_peer_name_matches_certificate(conn))
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
{
close_SSL(conn);
return PGRES_POLLING_FAILED;
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
}
/* SSL handshake is complete */
return PGRES_POLLING_OK;
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
}
/*
* Close SSL connection.
*/
static void
2002-09-04 22:31:48 +02:00
close_SSL(PGconn *conn)
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
{
if (conn->ssl)
{
DECLARE_SIGPIPE_INFO(spinfo);
DISABLE_SIGPIPE(conn, spinfo, (void) 0);
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
SSL_shutdown(conn->ssl);
SSL_free(conn->ssl);
conn->ssl = NULL;
pqsecure_destroy();
/* We have to assume we got EPIPE */
REMEMBER_EPIPE(spinfo, true);
RESTORE_SIGPIPE(conn, spinfo);
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
}
if (conn->peer)
{
X509_free(conn->peer);
conn->peer = NULL;
}
#ifdef USE_SSL_ENGINE
if (conn->engine)
{
ENGINE_finish(conn->engine);
ENGINE_free(conn->engine);
conn->engine = NULL;
}
#endif
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
}
/*
* Obtain reason string for last SSL error
*
* Some caution is needed here since ERR_reason_error_string will
* return NULL if it doesn't recognize the error code. We don't
* want to return NULL ever.
*/
static char ssl_nomem[] = "out of memory allocating error description";
2004-08-29 07:07:03 +02:00
#define SSL_ERR_LEN 128
static char *
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
SSLerrmessage(void)
{
2002-09-04 22:31:48 +02:00
unsigned long errcode;
const char *errreason;
2004-08-29 07:07:03 +02:00
char *errbuf;
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
errbuf = malloc(SSL_ERR_LEN);
if (!errbuf)
return ssl_nomem;
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
errcode = ERR_get_error();
2004-08-29 07:07:03 +02:00
if (errcode == 0)
{
snprintf(errbuf, SSL_ERR_LEN, libpq_gettext("no SSL error reported"));
return errbuf;
}
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
errreason = ERR_reason_error_string(errcode);
2004-08-29 07:07:03 +02:00
if (errreason != NULL)
{
strlcpy(errbuf, errreason, SSL_ERR_LEN);
return errbuf;
}
snprintf(errbuf, SSL_ERR_LEN, libpq_gettext("SSL error code %lu"), errcode);
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
return errbuf;
}
static void
SSLerrfree(char *buf)
{
if (buf != ssl_nomem)
free(buf);
}
2004-08-29 07:07:03 +02:00
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
/*
* Return pointer to OpenSSL object.
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
*/
void *
UPDATED PATCH: Attached are a revised set of SSL patches. Many of these patches are motivated by security concerns, it's not just bug fixes. The key differences (from stock 7.2.1) are: *) almost all code that directly uses the OpenSSL library is in two new files, src/interfaces/libpq/fe-ssl.c src/backend/postmaster/be-ssl.c in the long run, it would be nice to merge these two files. *) the legacy code to read and write network data have been encapsulated into read_SSL() and write_SSL(). These functions should probably be renamed - they handle both SSL and non-SSL cases. the remaining code should eliminate the problems identified earlier, albeit not very cleanly. *) both front- and back-ends will send a SSL shutdown via the new close_SSL() function. This is necessary for sessions to work properly. (Sessions are not yet fully supported, but by cleanly closing the SSL connection instead of just sending a TCP FIN packet other SSL tools will be much happier.) *) The client certificate and key are now expected in a subdirectory of the user's home directory. Specifically, - the directory .postgresql must be owned by the user, and allow no access by 'group' or 'other.' - the file .postgresql/postgresql.crt must be a regular file owned by the user. - the file .postgresql/postgresql.key must be a regular file owned by the user, and allow no access by 'group' or 'other'. At the current time encrypted private keys are not supported. There should also be a way to support multiple client certs/keys. *) the front-end performs minimal validation of the back-end cert. Self-signed certs are permitted, but the common name *must* match the hostname used by the front-end. (The cert itself should always use a fully qualified domain name (FDQN) in its common name field.) This means that psql -h eris db will fail, but psql -h eris.example.com db will succeed. At the current time this must be an exact match; future patches may support any FQDN that resolves to the address returned by getpeername(2). Another common "problem" is expiring certs. For now, it may be a good idea to use a very-long-lived self-signed cert. As a compile-time option, the front-end can specify a file containing valid root certificates, but it is not yet required. *) the back-end performs minimal validation of the client cert. It allows self-signed certs. It checks for expiration. It supports a compile-time option specifying a file containing valid root certificates. *) both front- and back-ends default to TLSv1, not SSLv3/SSLv2. *) both front- and back-ends support DSA keys. DSA keys are moderately more expensive on startup, but many people consider them preferable than RSA keys. (E.g., SSH2 prefers DSA keys.) *) if /dev/urandom exists, both client and server will read 16k of randomization data from it. *) the server can read empheral DH parameters from the files $DataDir/dh512.pem $DataDir/dh1024.pem $DataDir/dh2048.pem $DataDir/dh4096.pem if none are provided, the server will default to hardcoded parameter files provided by the OpenSSL project. Remaining tasks: *) the select() clauses need to be revisited - the SSL abstraction layer may need to absorb more of the current code to avoid rare deadlock conditions. This also touches on a true solution to the pg_eof() problem. *) the SIGPIPE signal handler may need to be revisited. *) support encrypted private keys. *) sessions are not yet fully supported. (SSL sessions can span multiple "connections," and allow the client and server to avoid costly renegotiations.) *) makecert - a script that creates back-end certs. *) pgkeygen - a tool that creates front-end certs. *) the whole protocol issue, SASL, etc. *) certs are fully validated - valid root certs must be available. This is a hassle, but it means that you *can* trust the identity of the server. *) the client library can handle hardcoded root certificates, to avoid the need to copy these files. *) host name of server cert must resolve to IP address, or be a recognized alias. This is more liberal than the previous iteration. *) the number of bytes transferred is tracked, and the session key is periodically renegotiated. *) basic cert generation scripts (mkcert.sh, pgkeygen.sh). The configuration files have reasonable defaults for each type of use. Bear Giles
2002-06-14 06:23:17 +02:00
PQgetssl(PGconn *conn)
{
if (!conn)
return NULL;
return conn->ssl;
}
2005-10-15 04:49:52 +02:00
#else /* !USE_SSL */
void *
PQgetssl(PGconn *conn)
{
return NULL;
}
2002-09-04 22:31:48 +02:00
#endif /* USE_SSL */
#if defined(ENABLE_THREAD_SAFETY) && !defined(WIN32)
/*
2005-10-15 04:49:52 +02:00
* Block SIGPIPE for this thread. This prevents send()/write() from exiting
* the application.
*/
int
pq_block_sigpipe(sigset_t *osigset, bool *sigpipe_pending)
{
2005-10-15 04:49:52 +02:00
sigset_t sigpipe_sigset;
sigset_t sigset;
sigemptyset(&sigpipe_sigset);
sigaddset(&sigpipe_sigset, SIGPIPE);
/* Block SIGPIPE and save previous mask for later reset */
SOCK_ERRNO_SET(pthread_sigmask(SIG_BLOCK, &sigpipe_sigset, osigset));
if (SOCK_ERRNO)
return -1;
/* We can have a pending SIGPIPE only if it was blocked before */
if (sigismember(osigset, SIGPIPE))
{
/* Is there a pending SIGPIPE? */
if (sigpending(&sigset) != 0)
return -1;
2005-10-15 04:49:52 +02:00
if (sigismember(&sigset, SIGPIPE))
*sigpipe_pending = true;
else
*sigpipe_pending = false;
}
else
*sigpipe_pending = false;
2005-10-15 04:49:52 +02:00
return 0;
}
2005-10-15 04:49:52 +02:00
/*
* Discard any pending SIGPIPE and reset the signal mask.
*
* Note: we are effectively assuming here that the C library doesn't queue
2005-10-15 04:49:52 +02:00
* up multiple SIGPIPE events. If it did, then we'd accidentally leave
* ours in the queue when an event was already pending and we got another.
* As long as it doesn't queue multiple events, we're OK because the caller
* can't tell the difference.
*
* The caller should say got_epipe = FALSE if it is certain that it
* didn't get an EPIPE error; in that case we'll skip the clear operation
* and things are definitely OK, queuing or no. If it got one or might have
* gotten one, pass got_epipe = TRUE.
*
* We do not want this to change errno, since if it did that could lose
2005-10-15 04:49:52 +02:00
* the error code from a preceding send(). We essentially assume that if
* we were able to do pq_block_sigpipe(), this can't fail.
*/
void
pq_reset_sigpipe(sigset_t *osigset, bool sigpipe_pending, bool got_epipe)
{
2005-10-15 04:49:52 +02:00
int save_errno = SOCK_ERRNO;
int signo;
sigset_t sigset;
2004-08-29 07:07:03 +02:00
/* Clear SIGPIPE only if none was pending */
if (got_epipe && !sigpipe_pending)
{
if (sigpending(&sigset) == 0 &&
sigismember(&sigset, SIGPIPE))
{
2005-10-15 04:49:52 +02:00
sigset_t sigpipe_sigset;
sigemptyset(&sigpipe_sigset);
sigaddset(&sigpipe_sigset, SIGPIPE);
sigwait(&sigpipe_sigset, &signo);
}
}
2005-10-15 04:49:52 +02:00
/* Restore saved block mask */
pthread_sigmask(SIG_SETMASK, osigset, NULL);
SOCK_ERRNO_SET(save_errno);
}
#endif /* ENABLE_THREAD_SAFETY && !WIN32 */