summaryrefslogtreecommitdiffstats
path: root/base/message_pump_libevent.cc
Commit message (Collapse)AuthorAgeFilesLines
* POSIX: Add a macro for handling EINTR.agl@chromium.org2009-05-011-2/+3
| | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | | On POSIX systems, system calls can be interrupted by signals. In this case, they'll return EINTR, indicating that the system call needs to be restarted. (The situation is a little more complicated than this with SA_RESTART, but you can read man 7 signal if you like.) The short of it is that you need to catch EINTR and restart the call for these system calls: * read, readv, write, writev, ioctl * open() when dealing with a fifo * wait* * Anything socket based (send*, recv*, connect, accept etc) * flock and lock control with fcntl * mq_ functions which can block * futex * sem_wait (and timed wait) * pause, sigsuspend, sigtimedwait, sigwaitinfo * poll, epoll_wait, select and 'p' versions of the same * msgrcv, msgsnd, semop, semtimedop * close (although, on Linux, EINTR won't happen here) * any sleep functions (careful, you need to handle this are restart with different arguments) We've been a little sloppy with this until now. This patch adds a macro for dealing with this and corrects every case of these system calls (that I found). The macro is HANDLE_EINTR in base/eintr_wrapper.h. It's safe to include on Windows and is a no-op there. On POSIX, it uses GCC magic to return the correct type based on the expression and restarts the system call if it throws EINTR. And you can use it like: HANDLE_EINTR(close(fd)); Or: ssize_t bytes_read = HANDLE_EINTR(read(fd, buffer, len)); *BEWARE* that it will evaluate the argument multiple times, so this is not safe: HANDLE_EINTR(close(FireMissiles())); http://groups.google.com/group/chromium-dev/browse_thread/thread/41a35b2a457d73a0 http://codereview.chromium.org/100225 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@15102 0039d316-1c4b-4281-b951-d872f2087c98
* StopWatchingFileDescriptor needs to free struct event.dkegel@google.com2009-04-221-9/+32
| | | | | | | | | | | Also, there's no point in using scoped_ptr for event_ anymore, so removed that. Should fix http://crbug.com/10503 "Crash in network layer" Review URL: http://codereview.chromium.org/87045 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@14233 0039d316-1c4b-4281-b951-d872f2087c98
* Minor DCHECK cleanups in libevent.willchan@chromium.org2009-03-171-2/+3
| | | | | | | | DCHECK=>DCHECK_EQ where possible. Change a DCHECK not to fire on write() returning -1 && errno==EAGAIN, since this happens normally under heavy load. Review URL: http://codereview.chromium.org/42295 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@11919 0039d316-1c4b-4281-b951-d872f2087c98
* The dtor wasn't actually cleaning up the fds from the pipe, so close them downthomasvl@chromium.org2009-02-031-0/+4
| | | | | | | | during shutdown. (this w/ the zombies was what caused the bots to run out of fds) Review URL: http://codereview.chromium.org/20006 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@9066 0039d316-1c4b-4281-b951-d872f2087c98
* Followup for one missed review comment (sorting includes).thomasvl@chromium.org2009-01-201-1/+1
| | | | git-svn-id: svn://svn.chromium.org/chrome/trunk/src@8329 0039d316-1c4b-4281-b951-d872f2087c98
* Added more logging in debug builds for some failure passes, hopefully thisthomasvl@chromium.org2009-01-201-3/+10
| | | | | | | | will let us figure out why we see so mainly unittest failures on the mac but never seem to be able to repro locally. Review URL: http://codereview.chromium.org/18378 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@8316 0039d316-1c4b-4281-b951-d872f2087c98
* message_pump_libevent refactor:jeremy@chromium.org2008-12-151-57/+93
| | | | | | | | | | * Unify WatchSocket & WatchFileHandle. * Better encapsulate libevent. * Fix a bug with blocking writes in ipc_posix.cc Review URL: http://codereview.chromium.org/13757 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@7010 0039d316-1c4b-4281-b951-d872f2087c98
* Reverting 6911.ojan@google.com2008-12-121-77/+57
| | | | | | Review URL: http://codereview.chromium.org/14068 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@6916 0039d316-1c4b-4281-b951-d872f2087c98
* message_pump_libevent refactor:jeremy@chromium.org2008-12-121-57/+77
| | | | | | | | | | * Unify WatchSocket & WatchFileHandle. * Better encapsulate libevent. * Fix a bug with blocking writes in ipc_posix.cc Review URL: http://codereview.chromium.org/13757 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@6911 0039d316-1c4b-4281-b951-d872f2087c98
* Add FD watching support to message_pump_libevent.jeremy@chromium.org2008-12-081-4/+39
| | | | | | | | This CL goes along the same lines as the WatchSocket code, but we really need better encapsulation for this class in generel. Review URL: http://codereview.chromium.org/13129 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@6513 0039d316-1c4b-4281-b951-d872f2087c98
* Move Time, TimeDelta and TimeTicks into namespace base.dsh@google.com2008-10-271-1/+0
| | | | | | Review URL: http://codereview.chromium.org/7995 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@4022 0039d316-1c4b-4281-b951-d872f2087c98
* Provide autorelease pool management on the Mac in all types of message pumps.mark@chromium.org2008-09-251-2/+5
| | | | | | | Fix a few unit tests to use autorelease pools as needed. Review URL: http://codereview.chromium.org/3805 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@2603 0039d316-1c4b-4281-b951-d872f2087c98
* Use libevent, second try. Changes this time:dkegel@google.com2008-09-181-0/+179
| | | | | | | | | - remove bogus include of base/completion_callback.h - add DEPS rules to allow including third_party/libevent Review URL: http://codereview.chromium.org/2964 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@2371 0039d316-1c4b-4281-b951-d872f2087c98
* Revert "Make tcp_client_socket_unittest pass on Linux."evanm@google.com2008-09-181-179/+0
| | | | | | | This reverts commit d63170654d935f5a8ff6517bfacb60893f186a6b. git-svn-id: svn://svn.chromium.org/chrome/trunk/src@2347 0039d316-1c4b-4281-b951-d872f2087c98
* Make tcp_client_socket_unittest pass on Linux.dkegel@google.com2008-09-181-0/+179
Requires another changeset that puts libevent in third_party; I'll upload that next. This is not the final word; it makes too many syscalls per read. But it's a start. Review URL: http://codereview.chromium.org/3202 git-svn-id: svn://svn.chromium.org/chrome/trunk/src@2346 0039d316-1c4b-4281-b951-d872f2087c98