If one host in a multi-host connection string times out, move on to
the next specified host instead of giving up entirely.
Takayuki Tsunakawa, reviewed by Michael Paquier. I added
a minor adjustment to the documentation.
Discussion: http://postgr.es/m/0A3221C70F24FB45833433255569204D1F6F42F5@G01JPEXMBYT05
poll.h is mandated by Single Unix Spec v2, the usual baseline for
postgres on unix. None of the unixoid buildfarms animals has
sys/poll.h but not poll.h. Therefore there's not much point to test
for sys/poll.h's existence and include it optionally.
Author: Andres Freund, per suggestion from Tom Lane
Discussion: https://postgr.es/m/20505.1492723662@sss.pgh.pa.us
In commit 210eb9b743 I centralized libpq's logic for closing down
the backend communication socket, and made the new pqDropConnection
routine always reset the I/O buffers to empty. Many of the call sites
previously had not had such code, and while that amounted to an oversight
in some cases, there was one place where it was intentional and necessary
*not* to flush the input buffer: pqReadData should never cause that to
happen, since we probably still want to process whatever data we read.
This is the true cause of the problem Robert was attempting to fix in
c3e7c24a1d, namely that libpq no longer reported the backend's final
ERROR message before reporting "server closed the connection unexpectedly".
But that only accidentally fixed it, by invoking parseInput before the
input buffer got flushed; and very likely there are timing scenarios
where we'd still lose the message before processing it.
To fix, pass a flag to pqDropConnection to tell it whether to flush the
input buffer or not. On review I think flushing is actually correct for
every other call site.
Back-patch to 9.3 where the problem was introduced. In HEAD, also improve
the comments added by c3e7c24a1d.
If libpq output buffer is full, pqSendSome() function tries to drain any
incoming data. This avoids deadlock, if the server e.g. sends a lot of
NOTICE messages, and blocks until we read them. However, pqSendSome() only
did that in blocking mode. In non-blocking mode, the deadlock could still
happen.
To fix, take a two-pronged approach:
1. Change the documentation to instruct that when PQflush() returns 1, you
should wait for both read- and write-ready, and call PQconsumeInput() if it
becomes read-ready. That fixes the deadlock, but applications are not going
to change overnight.
2. In pqSendSome(), drain the input buffer before returning 1. This
alleviates the problem for applications that only wait for write-ready. In
particular, a slow but steady stream of NOTICE messages during COPY FROM
STDIN will no longer cause a deadlock. The risk remains that the server
attempts to send a large burst of data and fills its output buffer, and at
the same time the client also sends enough data to fill its output buffer.
The application will deadlock if it goes to sleep, waiting for the socket
to become write-ready, before the server's data arrives. In practice,
NOTICE messages and such that the server might be sending are usually
short, so it's highly unlikely that the server would fill its output buffer
so quickly.
Backpatch to all supported versions.
The EOF-detection logic in pqReadData was a bit confused about who should
set up the error message in case the kernel gives us read-ready-but-no-data
rather than ECONNRESET or some other explicit error condition. Since the
whole point of this situation is that the lower-level functions don't know
there's anything wrong, pqReadData itself must set up the message. But
keep the assumption that if an errno was reported, a message was set up at
lower levels.
Per bug #11712 from Marko Tiikkaja. It's been like this for a very long
time, so back-patch to all supported branches.
This patch makes libpq check the server's hostname against DNS names listed
in the X509 subjectAltName extension field in the server certificate. This
allows the same certificate to be used for multiple domain names. If there
are no SANs in the certificate, the Common Name field is used, like before
this patch. If both are given, the Common Name is ignored. That is a bit
surprising, but that's the behavior mandated by the relevant RFCs, and it's
also what the common web browsers do.
This also adds a libpq_ngettext helper macro to allow plural messages to be
translated in libpq. Apparently this happened to be the first plural message
in libpq, so it was not needed before.
Alexey Klyukin, with some kibitzing by me.
This refactoring is in preparation for adding support for other SSL
implementations, with no user-visible effects. There are now two #defines,
USE_OPENSSL which is defined when building with OpenSSL, and USE_SSL which
is defined when building with any SSL implementation. Currently, OpenSSL is
the only implementation so the two #defines go together, but USE_SSL is
supposed to be used for implementation-independent code.
The libpq SSL code is changed to use a custom BIO, which does all the raw
I/O, like we've been doing in the backend for a long time. That makes it
possible to use MSG_NOSIGNAL to block SIGPIPE when using SSL, which avoids
a couple of syscall for each send(). Probably doesn't make much performance
difference in practice - the SSL encryption is expensive enough to mask the
effect - but it was a natural result of this refactoring.
Based on a patch by Martijn van Oosterhout from 2006. Briefly reviewed by
Alvaro Herrera, Andreas Karlsson, Jeff Janes.
Based on the old comment, it took me a while to figure out what the
problem was. The importnat detail is that SSL_read() can return WANT_READ
even though some raw data was received from the socket.
If the server sends a long stream of data, and the server + network are
consistently fast enough to force the recv() loop in pqReadData() to
iterate until libpq's input buffer is full, then upon processing the last
incomplete message in each bufferload we'd usually double the buffer size,
due to supposing that we didn't have enough room in the buffer to finish
collecting that message. After filling the newly-enlarged buffer, the
cycle repeats, eventually resulting in an out-of-memory situation (which
would be reported misleadingly as "lost synchronization with server").
Of course, we should not enlarge the buffer unless we still need room
after discarding already-processed messages.
This bug dates back quite a long time: pqParseInput3 has had the behavior
since perhaps 2003, getCopyDataMessage at least since commit 70066eb1a1
in 2008. Probably the reason it's not been isolated before is that in
common environments the recv() loop would always be faster than the server
(if on the same machine) or faster than the network (if not); or at least
it wouldn't be slower consistently enough to let the buffer ramp up to a
problematic size. The reported cases involve Windows, which perhaps has
different timing behavior than other platforms.
Per bug #7914 from Shin-ichi Morita, though this is different from his
proposed solution. Back-patch to all supported branches.
Previously, 'int' was used for socket values in libpq, but socket values
are unsigned on Windows. This is a style correction.
Initial patch and previous PGINVALID_SOCKET initial patch by Joel
Jacobson, modified by me
Report from PVS-Studio
In pqSendSome, if the connection is already closed at entry, discard any
queued output data before returning. There is no possibility of ever
sending the data, and anyway this corresponds to what we'd do if we'd
detected a hard error while trying to send(). This avoids possible
indefinite bloat of the output buffer if the application keeps trying
to send data (or even just keeps trying to do PQputCopyEnd, as psql
indeed will).
Because PQputCopyEnd won't transition out of PGASYNC_COPY_IN state
until it's successfully queued the COPY END message, and pqPutMsgEnd
doesn't distinguish a queuing failure from a pqSendSome failure,
this omission allowed an infinite loop in psql if the connection closure
occurred when we had at least 8K queued to send. It might be worth
refactoring so that we can make that distinction, but for the moment
the other changes made here seem to offer adequate defenses.
To guard against other variants of this scenario, do not allow
PQgetResult to return a PGRES_COPY_XXX result if the connection is
already known dead. Make sure it returns PGRES_FATAL_ERROR instead.
Per report from Stephen Frost. Back-patch to all active branches.
We had two copies of this function in the backend and libpq, which was
already pretty bogus, but it turns out that we need it in some other
programs that don't use libpq (such as pg_test_fsync). So put it where
it probably should have been all along. The signal-mask-initialization
support in src/backend/libpq/pqsignal.c stays where it is, though, since
we only need that in the backend.
Create an internal function pqDropConnection that does the physical socket
close and cleans up closely-associated state. This removes a bunch of ad
hoc, not always consistent closure code. The ulterior motive is to have a
single place to wait for a spawned child backend to exit, but this seems
like good cleanup even if that never happens.
I went back and forth on whether to include "conn->status = CONNECTION_BAD"
in pqDropConnection's actions, but for the moment decided not to. Only a
minority of the call sites actually want that, and in any case it's
arguable that conn->status is slightly higher-level state, and thus not
part of this function's purview.
Traditionally libpq has collected an entire query result before passing
it back to the application. That provides a simple and transactional API,
but it's pretty inefficient for large result sets. This patch allows the
application to process each row on-the-fly instead of accumulating the
rows into the PGresult. Error recovery becomes a bit more complex, but
often that tradeoff is well worth making.
Kyotaro Horiguchi, reviewed by Marko Kreen and Tom Lane
On balance, the need to cover this case changes my mind in favor of pushing
all error-message generation duties into the two fe-secure.c routines.
So do it that way.
In many cases, pqsecure_read/pqsecure_write set up useful error messages,
which were then overwritten with useless ones by their callers. Fix this
by defining the responsibility to set an error message to be entirely that
of the lower-level function when using SSL.
Back-patch to 8.3; the code is too different in 8.2 to be worth the
trouble.
This function is like the PQserverVersion() function except
it returns the version of libpq, making it possible for a client
program or driver to determine which version of libpq is in
use at runtime, and not just at link time.
Suggested by Harald Armin Massa and several others.
My initial impression that glibc was measuring the precision in characters
(which is what the Linux man page says it does) was incorrect. It does take
the precision to be in bytes, but it also tries to truncate the string at a
character boundary. The bottom line remains the same: it will mess up
if the string is not in the encoding it expects, so we need to avoid %.*s
anytime there's a significant risk of that. Previous code changes are still
good, but adjust the comments to reflect this knowledge. Per research by
Hernan Gonzalez.
Depending on which spec you read, field widths and precisions in %s may be
counted either in bytes or characters. Our code was assuming bytes, which
is wrong at least for glibc's implementation, and in any case libc might
have a different idea of the prevailing encoding than we do. Hence, for
portable results we must avoid using anything more complex than just "%s"
unless the string to be printed is known to be all-ASCII.
This patch fixes the cases I could find, including the psql formatting
failure reported by Hernan Gonzalez. In HEAD only, I also added comments
to some places where it appears safe to continue using "%.*s".
after each other (since we already add a newline on each, this makes them
multiline).
Previously a new error would just overwrite the old one, so for example any
error caused when trying to connect with SSL enabled would be overwritten
by the error message form the non-SSL connection when using sslmode=prefer.
Per Microsoft knowledge base article Q201213, early versions of
Windows fail when we do this. Later versions of Windows appear
to have a higher limit than 64Kb, but do still fail on large
sends, so we unconditionally limit it for all versions.
Patch from Tom Lane.
we are on a 64-bit machine (ie, size_t is wider than int) and someone passes
in a query string that approaches or exceeds INT_MAX bytes. Also, just for
paranoia's sake, guard against similar overflows in sizing the input buffer.
The backend will not in the foreseeable future be prepared to send or receive
strings exceeding 1GB, so I didn't take the more invasive step of switching
all the buffer index variables from int to size_t; though someday we might
want to do that.
I have a suspicion that this is not the only such bug in libpq, but this
fix is enough to take care of the crash reported by Francisco Reyes.
renumbering of encoding IDs done between 8.2 and 8.3 turns out to break 8.2
initdb and psql if they are run with an 8.3beta1 libpq.so. For the moment
we can rearrange the order of enum pg_enc to keep the same number for
everything except PG_JOHAB, which isn't a problem since there are no direct
references to it in the 8.2 programs anyway. (This does force initdb
unfortunately.)
Going forward, we want to fix things so that encoding IDs can be changed
without an ABI break, and this commit includes the changes needed to allow
libpq's encoding IDs to be treated as fully independent of the backend's.
The main issue is that libpq clients should not include pg_wchar.h or
otherwise assume they know the specific values of libpq's encoding IDs,
since they might encounter version skew between pg_wchar.h and the libpq.so
they are using. To fix, have libpq officially export functions needed for
encoding name<=>ID conversion and validity checking; it was doing this
anyway unofficially.
It's still the case that we can't renumber backend encoding IDs until the
next bump in libpq's major version number, since doing so will break the
8.2-era client programs. However the code is now prepared to avoid this
type of problem in future.
Note that initdb is no longer a libpq client: we just pull in the two
source files we need directly. The patch also fixes a few places that
were being sloppy about checking for an unrecognized encoding name.
o remove many WIN32_CLIENT_ONLY defines
o add WIN32_ONLY_COMPILER define
o add 3rd argument to open() for portability
o add include/port/win32_msvc directory for
system includes
Magnus Hagander