| Commit message (Collapse) | Author | Age | Files | Lines |
|
|
|
|
|
|
|
|
|
|
|
| |
* We need to disable inheritance in the case of success like in the
cases for stdout and stdin.
Fixes the same issue fixed for unix in r783398
Submitted by: rpluem
Reviewed by: rpluem
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/branches/1.4.x@783725 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
| |
* We need to disable inheritance in the case of success like in the
cases for stdout and stdin.
Submitted by: rpluem
Reviewed by: rpluem
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/branches/1.4.x@783399 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@720054 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@649208 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@617375 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
simplification, it fixes a compile error in the call to
pthread_yield() on z/OS.
Submitted by: David Jones
I modified it slightly to use AC_DEFINE() as suggested
by jorton.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@596402 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@590849 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
| |
We still need to restore the std handles if "no file" (filedes == -1)
is passed to the child.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@590848 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
| |
I haven't yet verified that the code works but this is a step in the right
direction.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@590037 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@585260 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
These may need massaging and do need review by their respective
communities.
Note that someone from the OS2 community needs to ping me with
resolving the missing apr_arch_inherit.h mess; this should be
very easy to translate into
DosSetFHState(handle, OPEN_FLAGS_NOINHERIT);
bits, but to more thoroughly resolve the issue, we should take
it a step further and consider the NT implementation which
toggles inheritance on only for handles as they hit proc_create,
so that you don't have cross-process handle leakage into the
wrong processes.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@584928 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
| |
of the pipe to uninherited. Let it be closed upon cleanup_for_exec.
The later dup2() for the parent pipe does not automagically become inherited
again, and later dup()'s are never inherited by default.
There's no longer an explicit need to close the parent-end in proc_create
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@584570 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
inherited. Solve this by setting the duplicated handle to inherit.
once finished with the fork(), now that we don't waste pipe creation
resources on a single handle, watch out for closing the parent handle
inside the child.
in fact I believe that toggling parent_* handles apr_file_inherit_unset
way back in apr_procattr_io_set / apr_procattr_child_*_set would be
more efficient; comments?
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@584569 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
First; we absolutely do NOT want to waste our time creating a pipe,
when the caller has their own file descriptors all set up to give to
the child process (and use itself). We can also presume a single
ended pipe is about as interesting as the sound of one hand clapping.
Create the pipe only when we don't already have any child/parent pipes
set up, and when the caller passes no files for us to use. Otherwise,
we simply dup for our own use rather than dup2.
Second; we absolutely cannot dup2 into the static 'no_file' special fd,
so we'll guard against this and also dup, instead, for this case.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@584500 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
| |
apr_file_pipe_create_ex() for win32, many lines are redundant.
Thanks to Eric for picking up on this mess in the first place!
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@584496 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
| |
Submitted by: Lucian Adrian Grijincu <lucian.grijincu gmail.com>
Reviewed by: rpluem
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@584411 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
| |
appropriate based on the input parameters.
PR: 43522
Submitted by: Eric Covener <covener gmail.com>
Reviewed by: trawick
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@583421 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
| |
(we aren't trying to close our child_fd's here).
Submitted by: David Glasser <glasser@davidglasser.net>
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@581089 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
| |
for pointing this out to me on #irc.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@581042 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
| |
These can be reduced further and still remain legible.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@580632 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@580591 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
| |
if we use apr_procattr_stderr_set() it will not close out the
previous handle parked there by _io_set(). But it also does
not attempt to touch the _io_set() no_file STATIC apr_file_t's
so there is nothing to otherwise fix here immediately.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@580515 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
| |
to cause the specified streams to be closed to the child process,
when the caller has chosen that flag via apr_procattr_io_set().
ALSO; solve a serious flaw where we attempted to dup2 to a non existant
file if the user had not already called apr_procattr_io_set()!
The Unix implementation.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@580486 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
| |
to cause the specified streams to be closed to the child process,
when the caller has chosen that flag via apr_procattr_io_set().
This is the nonportable flavor targeting 1.2.12; unix 1.3.0 specific
commit to follow.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@580484 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
| |
caused StdOutput to never be inherited (something not noticed,
nor particularly interesting in httpd, which is how I missed it).
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@572457 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
| |
thankfully.
Submitted by: Steven Narin <steven.narin googlemail.com>
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@571712 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@569890 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Solve win32 inherited pipe leaks by leveraging OS2 port's solution.
Mutex the pipe manipulation on WinNT+++ alone (not WinCE, nor 9x)
so that we toggle the inherited state of the stdin/out/err pipes.
This is only possible on NT, because in CE/9x it would involve
replacing the pipe handles all over the place as there is no toggle.
This CRITICAL_SECTION pipe is incredibly fast in the mainline case,
and only introduces contention in the threaded server after startup
(for cgi, etc). Not unlike an in-process cgid.
So, leave WinCE alone for now, since it doesn't follow the stdio model,
and leave Win9x alone for good, as nearly abandoned.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@569882 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
| |
stdin/out/err if some of these handles are left unset, while one
or two of them are set with apr_procattr_io_set() (with one or more
APR_NO_PIPE's) and/or apr_procattr_child_XXX_set().
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@568818 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
| |
".bat" or ".cmd" and falls through to the general case. There is no
COMSPEC on WinCE (no getenv either) and no command line shell.
PR: 39869
Submitted by: Curt Arnold <carnold apache.org>
Reviewed by: Davi Arnaut
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@543324 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
| |
PR: 33490/attachment 18106
Submitted by: Curt Arnold <carnold apache.org>
Reviewed by: Davi Arnaut <davi haxent.com.br>
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@543304 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
| |
PR: 39868
Submitted by: Curt Arnold <carnold apache.org>
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@543279 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
| |
Submitted by: David Jones <oscaremma gmail.com>
Reviewed by: trawick
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@501150 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
| |
Implement apr_threadkey_private destructors on WIN32
instead silently ignoring them.
and reviewed by wrowe, brane, jerenkrantz to need a more
comprehensive solution, e.g. Windows PE destructor entry
points, that does -not- break svn.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@478137 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
| |
executable-but-unreadable programs to be executed even if the errchk
mode is active.
PR: 32498
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@454787 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
| |
instead silently ignoring them.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@434327 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@428317 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@428313 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
| |
Originally titled: [patch 10/17] include path prefix win32/
Submitted by: John Mark Vandenberg
Reviewed by: Will Rowe (concept), Justin Erenkrantz
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@421349 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Submitted by: Keisuke Nishida <keisuke.nishida gmail.com>
Reviewed by: rooneg, Henry Jen <henryjen ztune.net>
* configure.in: Look for pthread_yield and sched_yield.
* include/arch/unix/apr_arch_threadproc.h: Include sched.h if it's present.
* threadproc/unix/thread.c
(apr_thread_yield): Actually do something in here...
* CHANGES: Note change.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@418351 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
| |
(apr_proc_detach): Remove tabs that were screwing up the indenting.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@396070 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
| |
Darwin zombie-avoidance hack on NetBSD too.
PR: 36750
Submitted by: Todd Vierling <tv pobox.com>
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@391580 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
| |
Noop in 32 bit, error in 64 bit builds.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@368551 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* threadproc/win32/proc.c
(apr_proc_create): Pass empty strings into CreateProcess instead of
filtering them out.
* CHANGES: Note change.
Submitted by: Philip Martin <philip codematters.co.uk>
Reviewed by: wrowe, rooneg
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@355792 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
| |
value (fixes gcc warning with modern glibc and -D_FORTIFY_SOURCE=2).
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@349408 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
| |
resolving a theoretical problem with standards compliance
and an actual problem with Purify.
Reviewed by: Joe Orton
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@312607 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
|
| |
prevent atexit-registered functions from being run in a failing child,
and e.g. flushing stdio buffers.
PR: 30913
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@239687 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
| |
the .exe extension is not given. Was trying to close an apr_file_t that had
failed to open, causing a segfault.
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@208879 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
|
|
|
|
|
| |
apr_procattr_child_out_set, apr_procattr_child_err_set): Add error
checking: fixes segfaults in applications when pipe() or dup2() fail
(e.g. due to ulimit settings).
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@170395 13f79535-47bb-0310-9956-ffa450edef68
|
|
|
|
| |
git-svn-id: http://svn.apache.org/repos/asf/apr/apr/trunk@151412 13f79535-47bb-0310-9956-ffa450edef68
|