IF YOU WOULD LIKE TO GET AN ACCOUNT, please write an
email to Administrator. User accounts are meant only to access repo
and report issues and/or generate pull requests.
This is a purpose-specific Git hosting for
BaseALT
projects. Thank you for your understanding!
Только зарегистрированные пользователи имеют доступ к сервису!
Для получения аккаунта, обратитесь к администратору.
On arm platforms incrementing a variable is not
an atomic operation, so may be interrupted by
signal processing (if a signal interrupts another
signal handler).
Use compiler built-ins to make this atomic.
__sync_fetch_and_add() works on gcc, llvm,
IBM xlC on AIX, and Intel icc (10.1 and
above).
atomic_add_32() works on Oracle Solaris.
Based on an inital patch from kamei@osstech.co.jp.
Bug #10640 - smbd is not responding - tevent_common_signal_handler() increments non-atomic variables
https://bugzilla.samba.org/show_bug.cgi?id=10640
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Volker Lendecke <Volker.Lendecke@SerNet.DE>
This way the caller can change use the supported limits without using hardcoded
values.
Bug: https://bugzilla.samba.org/show_bug.cgi?id=10214
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
There are some existing callers which assume the old
SA_INFO_QUEUE_COUNT 100 value.
256 should give room for the future.
Bug: https://bugzilla.samba.org/show_bug.cgi?id=10214
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Volker Lendecke <vl@samba.org>
It makes adding/removing the first/last sigevents a bit more expensive, but it
will fix tevent_loop_wait not finishing when one signal event was added and
removed.
Bug: https://bugzilla.samba.org/show_bug.cgi?id=10012
Signed-off-by: Volker Lendecke <vl@samba.org>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Reviewed-by: Andreas Schneider <asn@samba.org>
Signed-off-by: Volker Lendecke <vl@samba.org>
Reviewed-by: Michael Adam <obnox@samba.org>
Autobuild-User(master): Volker Lendecke <vl@samba.org>
Autobuild-Date(master): Fri Jun 21 19:57:06 CEST 2013 on sn-devel-104
Mask off signals the correct way from the signal handler.
Signed-off-by: Jeremy Allison <jra@samba.org>
Reviewed-by: Stefan Metzmacher <metze@samba.org>
Autobuild-User(master): Stefan Metzmacher <metze@samba.org>
Autobuild-Date(master): Tue Jan 15 12:13:43 CET 2013 on sn-devel-104
This fixes an uninitialized read introduced by my fix for the tevent_signal
destructors. From looking at the code you might believe that this kicks in only
when talloc failed. But with -O3 I do see it in normal operations.
Sorry for that.
Autobuild-User: Volker Lendecke <vl@samba.org>
Autobuild-Date: Wed Feb 15 17:58:37 CET 2012 on sn-devel-104
The results of some read(2) and write(2) calls are assigned into a
variable that is never used. Presumably this used to avoid compiler
warnings or similar.
However, from (approximately) GCC 4.6 this produces some warnings:
[ 609/3910] Compiling lib/tevent/tevent_signal.c
../lib/tevent/tevent_signal.c: In function ‘tevent_common_signal_handler’:
../lib/tevent/tevent_signal.c:85:10: warning: variable ‘res’ set but not used [-Wunused-but-set-variable]
../lib/tevent/tevent_signal.c: In function ‘signal_pipe_handler’:
../lib/tevent/tevent_signal.c:183:10: warning: variable ‘res’ set but not used [-Wunused-but-set-variable]
The simplest thing to do is remove the variables and cast the function
return to void. There is already a comment above each call.
Signed-off-by: Martin Schwenke <martin@meltin.net>
Signed-off-by: Stefan Metzmacher <metze@samba.org>
Make SA_RESETHAND conditional on its existance.
Autobuild-User: Jeremy Allison <jra@samba.org>
Autobuild-Date: Mon Aug 1 22:03:45 CEST 2011 on sn-devel-104
Revert 23abcd2318c69753aa2a144e1dc0f9cf9efdb705 and fix logic bug.
The current code loops through the event contexts, when it sees a different
one, it notifies the current one (ev) and updates ev to point to the new one.
This is dumb, because:
(1) ev starts as NULL, so this code crashes, and
(2) The final context will not be notified.
The correct fix for this is to update ev to the new one, then notify it.
Volker's fix works because we currently always have one event context.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
SA_INFO_QUEUE_COUNT *MUST* be a power of 2, in order for the ring buffer
wrap to work correctly at the 32 bit boundary. Thanks to Petr
Vandrovec <petr@vandrovec.name> for this.
When the first signal arrives, tevent_common_signal_handler() crashed: "ev" is
initialized to NULL, so the first "write(ev->pipe_fds[1], &c, 1);" dereferences
NULL.
Rusty, Tridge, please check. Also, can you tell me a bit more about the
environment you tested this in? I'd be curious to see where this survived.
Thanks,
Volker
I don't know if this is a problem in real life.
The code assumes there's only one tevent_context; all signals will notify
the first event context. That's counter-intuitive if you ever use more
than one, and there's nothing else in this code which prevents it AFAICT.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
We carefully preserve the old signal handler, but we replace it before
we've set up everything; in particular, if we fail setting up the
pipe_hack we could write a NUL char to stdout (fd 0), instead of
calling the old signal handler.
Replace the signal handler as the very last thing we do.
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
To be completely honest, I don't quite know whether to laugh or cry at
this one:
1 + (0xFFFFFFFF & ~(s.seen - s.count))
== 1 + (~(s.seen - s.count)) # s.seen, s.count are uint32_t
== s.count - s.seen # -A == ~A + 1
Signed-off-by: Rusty Russell <rusty@rustcorp.com.au>
Based on a patch submitted by Petr Vandrovec <petr@vandrovec.name>.
Multiple pending signals with siginfo_t's weren't being handled correctly
leading to smbd abort with kernel oplock signals.
Jeremy
The samba3 aio code requires lp_max_mux() with a default of 50
and the samba3 linux oplock code requires 100.
May we could make the size dynamic later.
metze
