|
de5428e7
|
2011-12-08T14:04:04
|
|
Merge remote-tracking branch 'fancycode/buffer_references'
Conflicts:
buffer.c
|
|
|
d4306028
|
2011-12-08T11:43:04
|
|
Merge remote-tracking branch 'origin/patches-2.0'
Conflicts:
configure.in
|
|
358c745e
|
2011-12-08T11:39:48
|
|
check for sysctl before we use it
Not all C libraries under Linux support the sysctl() func.
|
|
|
88a30ad8
|
2011-12-07T18:29:06
|
|
Split up extra-long AC_CHECK_FUNCS/HEADERS lines in configure.in
|
|
|
56c6afe2
|
2011-12-07T18:17:34
|
|
Merge branch '21_ifaddr'
|
|
|
7085a456
|
2011-12-07T18:16:32
|
|
Use getifaddrs to detect our interfaces if possible
The old scheme involved connected UDP sockets and getsockname(), and is
generally best avoied.
|
|
a8e5e2fc
|
2011-12-07T21:06:10
|
|
Changed use of refcounts to make sure referenced chains are freed in all cases.
|
|
|
27e22255
|
2011-12-07T13:04:35
|
|
Add evbuffer_copyout_from to copy data from the middle of a buffer
You could previously do this with evbuffer_peek() and some memcpys,
but it was a bit more work than most folks wanted to get into.
Closes sourceforge ticket 3108072
|
|
|
35c5c955
|
2011-12-07T11:49:52
|
|
Bypass event_add when using event_base_once() for a 0-sec timeout
Some people use event_base_once(EV_TIMEOUT) to make a callback get
called "immediately". But this is pretty roundabout: it uses the
timeout heap to immediately put the event onto the active queue, when
it could just use event_active. Additionally, it can lead to
surprising re-ordering behavior.
This patch changes event_base_once so it bypasses event_add() and
called event_active() directly on a pure-timeout event with an empty
timeout.
|
|
|
748a0d27
|
2011-12-06T13:38:40
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
19277763
|
2011-12-06T13:35:54
|
|
Fix a silly compilation error with the sun compiler
Apparently, other compilers were allowing "return free(x)" in a
function returning void.
|
|
|
9ce5053d
|
2011-12-06T11:02:12
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
|
11f36a5f
|
2011-12-05T15:02:27
|
|
Be absolutely sure to clear pncalls before leaving event_signal_closure
I thought we'd fixed the cases where this could come up, but
apparently having an event_base_break() happen while processing
signal events could get us in trouble.
Found by Remi Gacogne. Sourceforge issue 3451433 .
|
|
|
da45aa74
|
2011-12-02T01:48:14
|
|
fix windows compilation issues with new file segment code
Found by Linus Nordberg
|
|
|
f3b89dec
|
2011-11-24T12:31:50
|
|
Fix two issues in the allow_dirty_shutdown code.
First, it shouldn't crash when it's passed a non-ssl bufferevent.
Second, it should behave correctly when it gets a true argument
other than 1.
|
|
|
a44cd2b0
|
2011-11-24T12:24:38
|
|
Tweak allow_dirty_shutdown documentation
|
|
099d27df
|
2011-11-21T19:57:19
|
|
Allow users to set allow_dirty_shutdown
|
|
|
f7eb69ac
|
2011-11-21T19:24:50
|
|
bev_ssl: Be more specific in event callbacks. evhttp in particular gets confused without at least one of BEV_EVENT_{READING|WRITING}.
|
|
|
3596a31d
|
2011-11-18T15:35:33
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
|
ba4d6696
|
2011-11-18T15:34:37
|
|
Increment version to 2.0.16-stable-dev
|
|
|
5de3fa32
|
2011-11-18T15:27:06
|
|
Increment version to 2.0.16-stable
|
|
|
044c41de
|
2011-11-18T15:16:11
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
|
31c99bd2
|
2011-11-18T15:14:50
|
|
Revise 2.0.16-stable changelog
|
|
|
9ee9e943
|
2011-11-18T14:23:13
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
|
ebf82199
|
2011-11-17T17:42:45
|
|
add comment to new consider_reading code
|
|
2aa036fa
|
2011-11-17T11:59:41
|
|
Avoid spinning on OpenSSL reads
Previously, if some sender were generating data to read on an
OpenSSL connection as fast as we could process it, we could easily
wind up looping on an openssl do_read operation without ever
considering other sockets.
The difference between this and the original method in
consider_reading() is that it only loops for a single completed
*frame* instead of looping until fd is drained or an error condition
was triggered.
{Patch split out by nickm}
|
|
|
96c562fa
|
2011-11-17T11:54:07
|
|
Move SSL rate-limit enforcement into bytes_to_read()
|
|
|
a186e732
|
2011-11-17T11:45:49
|
|
Refactor amount-to-read calculations in buffervent_ssl consider_reading()
Split up consider_reading()'s conditional checks into another function
can_read() for simplicity sake.
{Split into a separate patch by nickm}
|
|
|
6660c9a3
|
2011-11-15T18:34:24
|
|
Revert "Avoid potential SSL read spinlocks"
This reverts commit fc52dbac87f4937f8306759506d6a2ad15ca244c.
|
|
|
a62346de
|
2011-11-15T18:33:50
|
|
Revert "openssl bufferevent has the same issue with writing as prior commit."
This reverts commit 7353663eb7c0b2a1caaaa5acd818515f156cf2ca.
|
|
|
fe9e831f
|
2011-11-15T17:22:45
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
|
34045c8f
|
2011-11-15T17:22:12
|
|
Add new entries to changelog, new credits to README
|
|
|
b17b0eec
|
2011-11-15T17:11:42
|
|
Fix DNS memleak checks when running with malloc-replacement/debugging disabled
|
|
|
f5b937e6
|
2011-11-14T18:22:08
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
|
e7874133
|
2011-11-14T17:33:02
|
|
Don't try to make notifiable event_base when no threading fns are configured
|
|
|
4e797f38
|
2011-11-14T17:32:22
|
|
Warn when unable to construct base because of failing make_base_notifiable
|
|
|
2ebc8c80
|
2011-11-14T11:55:37
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
|
7353663e
|
2011-11-14T10:57:15
|
|
openssl bufferevent has the same issue with writing as prior commit.
|
|
|
fc52dbac
|
2011-11-14T10:24:07
|
|
Avoid potential SSL read spinlocks
OpenSSL bufferevents with deferred callbacks enabled under high load will
spinlock in the function consider_reading(). This loop continues until all
data has been read.
Because of this condition; openssl bufferevents will never return back into
event_base_loop() until SSL_read has determined data is no longer ready.
As of yet I have not found a reason why this while loop exists, so this patch
just swaps out while for if.
If needed I can write same code which would trigger this effect; optionally
libevhtp has a test.c program which can be run with the following flags:
./test -s <keyfile.pem>
curl -vvvv -k -d@<HUGE_ASS_FILE> https://127.0.0.1:8081/
The return data will include the number of times the readcb got data and the
length of that read.
Without this patch, you are likely to see a small amount of "bytes read....",
otherwise the "bytes read..." return data should show much more reasonable
numbers.
|
|
|
9ae6e595
|
2011-11-14T11:45:17
|
|
Merge remote-tracking branch 'origin/patches-2.0'
Conflicts:
sample/dns-example.c
|
|
|
54142c94
|
2011-11-14T11:43:31
|
|
unit tests for EVBUFFER_EOL_NUL
|
|
d7a8b36e
|
2011-11-14T11:42:52
|
|
New EVBUFFER_EOL_NUL to read NUL-terminated strings from an evbuffer
|
|
|
27b5398f
|
2011-11-11T17:56:08
|
|
Tweaks, fixups, and comments on evbuffer_add_iovec
|
|
|
aaec5aca
|
2011-10-25T09:13:15
|
|
Added evbuffer_add_iovec and unit tests.
|
|
a3f320e8
|
2011-11-09T00:17:56
|
|
Improve win32 behavior of dns-sample.c codex
|
|
|
95af0433
|
2011-11-03T11:45:43
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
674bc6a2
|
2011-11-03T11:43:49
|
|
Fix typo in whatsnew-2.0.txt
|
|
|
dea055d0
|
2011-11-02T23:15:07
|
|
Merge remote-tracking branch 'github/20_evbuffer_remove_bug'
|
|
|
c882a959
|
2011-11-02T23:14:24
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
|
b18c04dd
|
2011-11-02T22:50:47
|
|
Use the free-trailing-chains function in evbuffer_insert_chain too
|
|
|
c37069cd
|
2011-11-02T16:09:15
|
|
Fix an evbuffer crash in evbuffer_remove_buffer()
Found by Greg Hazel.
|
|
7eb52eb8
|
2011-11-02T15:19:05
|
|
improve test to remove at least one buffer from src
|
|
|
90bd620f
|
2011-11-01T13:44:40
|
|
unit test for remove_buffer bug
|
|
|
00145918
|
2011-10-30T22:34:00
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
|
447b0bad
|
2011-10-30T22:32:18
|
|
Fix compile warning from saying event2/*.h inside a comment
Based on a patch by Adrian Chadd
|
|
|
0cb70e33
|
2011-10-26T10:17:21
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
1aee7183
|
2011-10-26T13:24:30
|
|
epoll: close fd on alloc fail at initialization
If the memory allocations fail then we free any other allocated
structures but don't close the file descriptor resulting in an leak of
fd's.
|
|
a2c48e3b
|
2011-10-24T01:31:55
|
|
evhttp: Add evhttp_foreach_bound_socket.
Applies the function specified in the first argument to all
evhttp_bound_sockets associated with a struct evhttp. The user
must not attempt to free or remove any connections, sockets or
listeners in the callback function.
|
|
|
356554c8
|
2011-10-25T11:43:01
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
|
3c824bd3
|
2011-10-24T13:18:09
|
|
Update copyright dates to 2011.
|
|
212533e4
|
2011-10-21T19:53:32
|
|
New event_base_update_cache_time() to set cached_tv to current time
This function is particularly useful for selectively increasing
the accuracy of the cached time value in 'base' during callbacks
that take a long time to execute.
This function has no effect if the base is currently not in its
event loop or if timeval caching is disabled via EVENT_BASE_FLAG_NO_CACHE_TIME.
|
|
|
c8953d1b
|
2011-10-14T17:16:03
|
|
Add argument checks to some memory functions in `event.c'.
Add a zero check to the function `event_mm_malloc_',
i.e. simply return NULL if the sz argument is zero.
On failure, set errno to ENOMEM and return NULL.
Add a zero check to the function `event_mm_calloc_',
i.e. simply return NULL if either argument is zero.
Also add an unsigned integer multiplication check, and if an integer
overflow would occur, set errno to ENOMEM and return NULL.
On failure, set errno to ENOMEM and return NULL.
Add a NULL check to the function `event_mm_strdup_',
i.e. set errno to EINVAL and return NULL.
Also add an unsigned integer addition check, and if an integer
overflow would occur, set errno to ENOMEM and return NULL.
If a memory allocation error occurs, again set errno to ENOMEM
and return NULL.
Add unit tests to `test/regress_util.c'.
|
|
|
724bfb56
|
2011-10-19T22:59:47
|
|
Add note about evhttp_send_reply_end to its doxygen
|
|
|
128c8d6c
|
2011-10-19T22:44:33
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
9e6a4efa
|
2011-10-19T17:44:17
|
|
More detailed message in case of libevent self-debugging failure.
|
|
|
5e42202d
|
2011-10-19T22:41:11
|
|
Style and comment tweaks for dns/leak* tests
|
|
|
f7841bfc
|
2011-10-18T17:49:40
|
|
Test for commit aff6ba1
|
|
|
fc1a2514
|
2011-10-19T22:20:15
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
|
21a08d68
|
2011-10-19T22:38:37
|
|
Empty DNS reply with OK status is another way to say NODATA.
Sometimes DNS reply has nothing but query section. It does not look like
error, so it should be treated as NODATA with TTL=0 as soon as there is
no SOA record to deduce negative TTL from.
|
|
|
d58c15ee
|
2011-10-19T22:36:12
|
|
Tests for 94fba5b and f72e8f6
|
|
|
95a8b87a
|
2011-10-17T21:48:23
|
|
fixed typo
|
|
|
ba24f616
|
2011-10-17T21:46:12
|
|
added comments to describe refcounting of multicase chains
|
|
|
e0b81ba3
|
2011-10-12T01:02:13
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
|
8589f6c4
|
2011-10-12T01:00:23
|
|
Bump version to 2.0.15-stable-dev
|
|
|
ab0f5f18
|
2011-10-12T00:57:55
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
|
c54c2358
|
2011-10-12T00:39:22
|
|
Credit a patch from a one-named user
|
|
|
78d80413
|
2011-10-12T00:22:54
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
|
4d9529fc
|
2011-10-12T00:22:17
|
|
Increment version to 2.0.15-stable
|
|
|
177f2868
|
2011-10-12T00:21:29
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
|
7c3e0612
|
2011-10-12T00:16:12
|
|
Add changelog for 2.0.15-stable
|
|
998c8138
|
2011-10-11T09:26:57
|
|
bufferevent: Add functions to set/get max_single_read/write values.
|
|
|
3c55b5ee
|
2011-10-11T09:50:57
|
|
Make evbase_priority_init() and evbase_get_npriorities() threadsafe
|
|
|
ee3a4ee8
|
2011-10-11T11:12:34
|
|
Add event_base_get_npriorities() function.
|
|
|
fed8f6e4
|
2011-10-10T11:34:26
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
|
ba5c27d4
|
2011-10-10T08:24:43
|
|
refer to non-deprecated evdns functions in comments
|
|
|
c6bbbf1b
|
2011-10-06T18:02:22
|
|
Make evbuffer_file_segment_types adaptable
Instead of having a file segment born as one type and stay that way
forever, let them start out unmapped, but map themselves as needed
if they need to get written out on a non-drains_to_fd evbuffer.
|
|
|
83588777
|
2011-10-06T15:21:55
|
|
Merge remote-tracking branch 'origin/patches-2.0'
Conflicts:
buffer.c
test/regress_buffer.c
|
|
|
706aa5f4
|
2011-10-06T15:11:50
|
|
Merge remote-tracking branch 'github/20_addfile_ssl' into patches-2.0
|
|
|
5760efb1
|
2011-10-05T15:03:17
|
|
Merge branch '21_largefile_support'
|
|
|
e0cae789
|
2011-10-05T11:26:52
|
|
Merge branch '21_evloop_noexit'
|
|
|
084e68f3
|
2010-12-01T21:53:08
|
|
New EVLOOP_NO_EXIT_ON_EMPTY option to keep looping even when no events are pending
This can be useful if you want to start an event loop and then add or
remove events to it from another thread.
|
|
|
9593a33f
|
2011-10-03T08:12:32
|
|
Allow evconnlistener to be created in disabled state.
|
|
|
b3bc77b6
|
2011-10-03T12:54:35
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
|
39c0cf7c
|
2011-10-03T12:49:02
|
|
Fix some "value never used" warnings with gcc 4.6.1
|
|
|
2b768479
|
2011-10-03T12:45:36
|
|
Make write-checking fixes use tt_fail_perror
|
|
|
c3b62fd7
|
2011-07-12T13:05:36
|
|
Fixed compiler warnings for unchecked read/write calls.
|
|
|
0ba0af9c
|
2011-09-29T09:30:04
|
|
Prefer mmap to sendfile unless a DRAINS_TO_FD flag is set. Allows add_file to work with SSL.
The sendfile() implementation for evbuffer_add_file is potentially more
efficient, but it has a problem: you can only use it to send bytes over
a socket using sendfile(). If you are writing bytes via SSL_send() or
via a filter, or if you need to be able to inspect your buffer, it
doesn't work.
As an easy fix, this patch disables the sendfile-based implementation of
evbuffer_add_file on an evbuffer unless the user sets a new
EVBUFFER_FLAG_DRAINS_TO_FD flag on that evbuffer, indicating that the
evbuffer will not be inspected, but only written out via
evbuffer_write(), evbuffer_write_atmost(), or drained with stuff like
evbuffer_drain() or evbuffer_add_buffer(). This flag is off by
default, except for evbuffers used for output on bufferevent_socket.
In the future, it could be interesting to make a best-effort file
segment implementation that tries to send via sendfile, but mmaps on
demand. That's too much complexity for a stable release series, though.
|
|
|
46974cd8
|
2011-09-28T09:23:41
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|
|
|
1ef1f684
|
2011-09-28T09:22:17
|
|
Make evbuffer callbacks get the right n_added value after evbuffer_add
Patch from Alex.
|
|
|
1ebe7955
|
2011-09-26T11:11:30
|
|
Merge remote-tracking branch 'origin/patches-2.0'
|