| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
| |
|
|\
| |
| | |
added initial .gitignore covering build results
|
| | |
|
|\ \
| |/
|/| |
Add help for systemd-based RH-like distro users
|
|/
|
|
|
|
|
|
|
|
| |
Recent RedHat-like distributions run distcc through systemd and
the configuration is not taken from this file, so add a comment that
put the poor user in the right direction.
This patch is RFC here because the downstream bug has not been
fixed and is going to be EOL'ed
See: https://bugzilla.redhat.com/show_bug.cgi?id=1187517
|
|\
| |
| | |
Adding missing README from svn to fix make install breakage
|
|/ |
|
|
|
| |
And some small cleanups.
|
| |
|
|
|
| |
Fixes issue 31
|
|
|
| |
Fixes issue 142
|
| |
|
| |
|
|
|
|
| |
An empty .gdbinit in the current directory is not effective at
avoiding ~/.gdbinit. Use the option -nh instead.
|
| |
|
|
|
|
|
|
|
| |
<http://code.google.com/p/distcc/issues/detail?id=139>:
add configure option --without-libiberty, to allow
building distcc without GPL3 license encumbrance.
|
|
|
|
|
|
| |
Passing the above option to configure should not cause ENABLE_RFC2553
to be defined.
Fixes issue 68
|
|
|
|
|
|
| |
Libiberty development files are now included in libiberty-dev, rather
than binutils-dev.
* INSTALL (QUICK SUMMARY): Update accordingly.
|
| |
|
|
|
|
|
|
|
|
|
|
|
| |
an exit status of 1 for compilation of a non-existing file.
This was causing problems with gcc 4.6.3 (and probably other
gcc versions) which was returning exit status 4 in that case.
I changed the test to instead use an existing file and
to expect exit status 0; we don't actually fail the compilation
if a bad log file is specified, instead we just log an error
message about the bad log file on stderr.
|
|
|
|
|
|
|
| |
since gcc 4.6.3 (and perhaps earlier) returns exit status 4.
Instead, just verify that the exit status from distcc is non-zero
and matches what gcc returned.
|
|
|
|
|
| |
fall back to compiling locally.
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
1. Add support for -isysroot and -arch.
2. Fix a really nasty corner case:
- CompressFiles::Compress uses os.makedirs to create a directory (foo)
for one compilation unit
- Another compilation unit does the equivalent of '#include
"foo/../bar.h"'
- The compilation unit uses no other file in foo
- We conclude that we don't have to create a dummy
foo/forcing_technique_271828
|
|
|
|
| |
* src/sendfile.c (sys_sendfile): Expose this function on GNU/Hurd,
which provides the same interface as Linux.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
* test/testdistcc.py: Remove stale and duplicate TODO items.
(_Touch): New function, acts like touch(1).
(Compile_c_Case): makeFile is not sufficient to reliably update
modification time on all systems (e.g. GNU/Hurd), so replace it with
calls to _Touch.
(Gdb_Case): Use _Touch.
(SyntaxError_Case): Look for pattern anywhere in the output using
re.search as warnings and other noise can be present.
* test/comfychair.py: Also set LC_ALL, which has a higher priority,
and make it so LANG and LC_ALL are set after saving the environment.
Clean up some stray import statements.
|
|
|
| |
Fixes issue 128
|
|
|
| |
Fixes issue 28
|
| |
|
|
|
| |
Fixes issue 117
|
|
|
| |
Fixes issue 113
|
|
|
| |
Fixes issue 127
|
|
|
| |
Fixes issue 85
|
|
|
|
|
|
|
|
| |
<http://code.google.com/p/distcc/issues/detail?id=114>
sent by Ying Chen <companycy@gmail.com>: ensure that lsdistcc.1.gz gets
included in the RPM package.
|
|
|
|
|
|
|
|
| |
<http://code.google.com/p/distcc/issues/detail?id=74>:
Return the host from backoff state, if local compilation
returned the same result as remote one.
|
|
|
|
|
|
| |
(using "make man-html").
Add HTML version of lsdistcc man page to index.html.
|
| |
|
| |
|
|
|
|
|
|
|
|
| |
Apply a patch from Debian that was forwarded upstream by mandyke.
The patch fixes a potential security problem in pump wrapper
(pump.in). If PYTHONPATH were originally unset or empty, the current working
directory would be added to PYTHONPATH, which is a security risk.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
include_server --no_force_dirs inhibits forcing_technique
The "forcing_technique" trick which handles the (rare) case of files
included as e.g. <foo/../file.h> causes a performance degradation:
The forcing files accumulate over the life of the include_server, making
distcc send more and more (useless) files as the build progresses.
While the forcing files are zero-length, the transmission filenames
themselves introduce a significant overhead for builds with many files
and directories.
This patch implements the include_server arg --no_force_dirs, which if
set switches off the "forcing_technique" scheme.
|
|
|
|
|
|
| |
Generalize the dcc_{r,x}_argv routines so they can be used to
transfer arbitrary array-of-strings tokens (useful for protocol
hacking experiments; no functional change).
|
|
|
|
|
|
|
| |
Fix for <http://bugs.debian.org/cgi-bin/bugreport.cgi?bug=638258>:
Distcc fails to build on hurd-i386 because it relies on PATH_MAX being
defined.
|
|
|
|
|
| |
handle header file names containing commas.
|
|
|
|
|
| |
:
lsdistcc.c compile fails: variable 'nready' set but not used [-Werror]
|
|
|
|
|
| |
pump should always report non-zero discrepancy count
|
|
|
|
|
|
|
|
| |
hardcoded 60 second backoff period should be tunable
Makes the backoff behavior tunable via a new env var DISTCC_BACKOFF_PERIOD.
The default backoff remains set for 60 seconds.
|
|
|
|
|
|
|
|
| |
"all servers busy" dcc_lock_pause is hardcoded at 1 sec, could be tunable.
Makes the pause time tunable via a new env var DISTCC_PAUSE_TIME_MSEC.
The default remains set for 1 second.
|
|
|
|
| |
<http://code.google.com/p/distcc/issues/detail?id=99>: include_server crash: local variable resolved referenced before assignment.
|
|
|
|
|
| |
build problem with -Werror due to ignoring return value from read().
|
|
|
|
|
| |
build problem with -Werror due to ignoring return value of 'chdir'.
|