Because checking in cmake breaks cross-compiling.
Introduced-in: 43eb56c7c738e3642f0981e3dd6ab9e082eec798.
Fixes: #482Fixes: #462
Refs: #475
v2: use waitid() with WNOWAIT
v3: use WNOWAIT only if it available in waitpid(), because not all netbsd
supports it
Before this patch we have one test.sh (well test-script.sh), and tooks
very long to run it sequentially, but they are pretty lightweight, so we
should run then in parallel.
Right now it fails because of regression for filtered openssl
bufferevent, and by it I mean ssl/bufferevent_filter_write_after_connect
test, and by fails - hang.
Regression-for: da52933550fd4736aa1c213b6de497e2ffc31e34 ("be_openssl:
don't call do_write() directly from outbuf_cb")
If system resolver (sync one) will respond too slow, then we can fail async
request and evdns will retransmit tham again, but evdns server will accept that
failed requets, so we will have not 2 requests but 4.
Reproduced on centos box sometimes.
As pointed in https://github.com/libevent/libevent/pull/417#issuecomment-267860738
"code is unsafe because in evutil_date_rfc1123() the pointer to the
automatic variable struct tm cur is used outside the scope it defined."
Checked with `clang -fsanitize=address -fsanitize-address-use-after-scope`
and test that call evutil_date_rfc1123() with tm==NULL
Tests:
- http/https_simple_dirty # not affected, since dirty is the default
- http/https_simple # affected
v2: fix compilation with -DEVENT__DISABLE_OPENSSL=ON
So firstly include our header (config.h) -- <evconfig-private.h>, and
only after it <sys/types.h> since latest has #ifdef guard, while our
config.h is not inteded for this.
And besides all this thing with LARGE_FILE is a abit awkward, since we
don't nefine _LP64/_LP32 anyway, and so we have next error actually (64bit VS
32bit):
==> solaris: In file included from ./util-internal.h:30:0,
==> solaris: from test/regress_ssl.c:49:
==> solaris: ./evconfig-private.h:29:0: warning: "_FILE_OFFSET_BITS" redefined
==> solaris: #define _FILE_OFFSET_BITS 64
==> solaris: ^
==> solaris: In file included from /usr/include/sys/types.h:17:0,
==> solaris: from test/regress_ssl.c:38:
==> solaris: /opt/csw/lib/gcc/i386-pc-solaris2.10/5.2.0/include-fixed/sys/feature_tests.h:196:0: note: this is the location of the previous definition
==> solaris: #define _FILE_OFFSET_BITS 32
==> solaris: ^
For cmake it commented in: 8b228e27f57300be61b57a41a2ec8666b726dc34
("Lot's of cmake updates")
==> netbsd: In file included from ../listener.c:57:0:
==> netbsd: ../util-internal.h:58:0: warning: "__func__" redefined [enabled by default]
==> netbsd: #define __func__ EVENT____func__
==> netbsd: ^
==> netbsd: In file included from /usr/include/amd64/types.h:39:0,
==> netbsd: from /usr/include/sys/types.h:45,
==> netbsd: from ../listener.c:30:
==> netbsd: /usr/include/sys/cdefs.h:394:0: note: this is the location of the previous definition
==> netbsd: #define __func__ __PRETTY_FUNCTION__
So move all of the declarations to the top of the offending function.
This patch includes both of issues (Fixes:), from @jeking3 and
@pprindeville
Fixes: #418Fixes: nmathewson/Libevent#136
Test that an event's callback is called if the fd is closed prior to being
polled for activity.
azat: make it run only for poll backend/method, and do not close fd
twice
Like in `sockaddr_in` structure in /usr/include/netinet/in.h
@azat: convert all other users (bench, compat, ..) and tweak message
Fixes: #178Fixes: #196
Refs: 6bf1ca78
Link: https://codereview.appspot.com/156040043/#msg4
According to solaris docs:
"One instance of a SIGCHLD signal is queued for each child process whose
status has changed. If waitpid() returns because the status of a child
process is available, and WNOWAIT was not specified in options, any pending
SIGCHLD signal associated with the process ID of that child process is
discarded. Any other pending SIGCHLD signals remain pending."
And interesting thing that it works if you add sleep(1) before waitpid(), and
also if you run with --verbose (some race or what).
But linux doesn't support WNOWAIT in waitpid() so add detection into
cmake/autotools.
Fixes: #387
Link: https://bugzilla.redhat.com/show_bug.cgi?id=840782
About this syncing pair:
- read endpoint, must be blocked, to make it a checkpoint or smth like this
- write endpoint, must be nonblocking, to avoid readcb hung
Refs: #387
Here's some fun. From `bufferevent.h`:
```
#define BEV_EVENT_READING 0x01 /**< error encountered while reading */
#define BEV_EVENT_WRITING 0x02 /**< error encountered while writing */
```
And from `event.h`:
```
/** Wait for a socket or FD to become readable */
#define EV_READ 0x02
/** Wait for a socket or FD to become writeable */
#define EV_WRITE 0x04
```
Library users have to be very careful to get this right; it turns out, the
library itself got this wrong in the `bufferevent_pair` code. It appears that
in most of the code, only `BEV_EVENT_FINISHED` will indicate whether it's read
or write; on error or timeout, it appears that "both" is assumed and not set in
the callback. I read through all the other places where `BEV_EVENT_FINISHED` is
passed to an event callback; it appears that the pair code is the only spot
that got it wrong.
azat: add TT_FORK to avoid breaking clean env, and rebase commit message
(copied from #359)
Fixes: #359
Using:
- evbuffer_add()
- evbuffer_add_buffer() -- the one that has problem
- evbuffer_add_reference() -- the only one that allows empty chains to be added
- evbuffer_remove_buffer()
