Movatterモバイル変換

[0]ホーム
URL:

Skip to content

Navigation Menu

Sign in
Appearance settings

Search code, repositories, users, issues, pull requests...

Provide feedback

We read every piece of feedback, and take your input very seriously.

Saved searches

Use saved searches to filter your results more quickly

 Sign up
Appearance settings

Commitde35a97

Browse files
committed
Handle WAIT_IO_COMPLETION return from WaitForMultipleObjectsEx().
This return code is possible wherever we pass bAlertable = TRUE; itarises when Windows caused the current thread to run an "I/O completionroutine" or an "asynchronous procedure call". PostgreSQL does notprovoke either of those Windows facilities, hence this bug remaininglargely unnoticed, but other local code might do so. Due to a shortageof complaints, no back-patch for now.Per report from Shiv Shivaraju Gowda, this bug can causePGSemaphoreLock() to PANIC. The bug can also cause select() to reporttimeout expiration too early, which might confuse pgstat_init() andCheckRADIUSAuth().
1 parente565ff7 commitde35a97

File tree

2 files changed

+37
-24
lines changed

2 files changed

+37
-24
lines changed

‎src/backend/port/win32/socket.c‎

Lines changed: 2 additions & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -623,7 +623,8 @@ pgwin32_select(int nfds, fd_set *readfds, fd_set *writefds, fd_set *exceptfds, c
623623
return0;
624624
}
625625

626-
if (r==WAIT_OBJECT_0+numevents)
626+
/* Signal-like events. */
627+
if (r==WAIT_OBJECT_0+numevents||r==WAIT_IO_COMPLETION)
627628
{
628629
pgwin32_dispatch_queued_signals();
629630
errno=EINTR;

‎src/backend/port/win32_sema.c‎

Lines changed: 35 additions & 23 deletions
Original file line numberDiff line numberDiff line change
@@ -118,8 +118,10 @@ PGSemaphoreReset(PGSemaphore sema)
118118
void
119119
PGSemaphoreLock(PGSemaphoresema,boolinterruptOK)
120120
{
121-
DWORDret;
122121
HANDLEwh[2];
122+
booldone= false;
123+
124+
ImmediateInterruptOK=interruptOK;
123125

124126
/*
125127
* Note: pgwin32_signal_event should be first to ensure that it will be
@@ -135,34 +137,44 @@ PGSemaphoreLock(PGSemaphore sema, bool interruptOK)
135137
* no hidden magic about whether the syscall will internally service a
136138
* signal --- we do that ourselves.
137139
*/
138-
do
140+
while (!done)
139141
{
140-
ImmediateInterruptOK=interruptOK;
141-
CHECK_FOR_INTERRUPTS();
142+
DWORDrc;
142143

143-
ret=WaitForMultipleObjectsEx(2,wh, FALSE,INFINITE, TRUE);
144+
CHECK_FOR_INTERRUPTS();
144145

145-
if (ret==WAIT_OBJECT_0)
146+
rc=WaitForMultipleObjectsEx(2,wh, FALSE,INFINITE, TRUE);
147+
switch (rc)
146148
{
147-
/* Signal event is set - we have a signal to deliver */
148-
pgwin32_dispatch_queued_signals();
149-
errno=EINTR;
149+
caseWAIT_OBJECT_0:
150+
/* Signal event is set - we have a signal to deliver */
151+
pgwin32_dispatch_queued_signals();
152+
break;
153+
caseWAIT_OBJECT_0+1:
154+
/* We got it! */
155+
done= true;
156+
break;
157+
caseWAIT_IO_COMPLETION:
158+
/*
159+
* The system interrupted the wait to execute an I/O
160+
* completion routine or asynchronous procedure call in this
161+
* thread. PostgreSQL does not provoke either of these, but
162+
* atypical loaded DLLs or even other processes might do so.
163+
* Now, resume waiting.
164+
*/
165+
break;
166+
caseWAIT_FAILED:
167+
ereport(FATAL,
168+
(errmsg("could not lock semaphore: error code %lu",
169+
GetLastError())));
170+
break;
171+
default:
172+
elog(FATAL,"unexpected return code from WaitForMultipleObjectsEx(): %lu",rc);
173+
break;
150174
}
151-
elseif (ret==WAIT_OBJECT_0+1)
152-
{
153-
/* We got it! */
154-
errno=0;
155-
}
156-
else
157-
/* Otherwise we are in trouble */
158-
errno=EIDRM;
159-
160-
ImmediateInterruptOK= false;
161-
}while (errno==EINTR);
175+
}
162176

163-
if (errno!=0)
164-
ereport(FATAL,
165-
(errmsg("could not lock semaphore: error code %lu",GetLastError())));
177+
ImmediateInterruptOK= false;
166178
}
167179

168180
/*

0 commit comments

Comments
 (0)
[8]ページ先頭
©2009-2025 Movatter.jp