minaco2 / distcc

Automatically exported from code.google.com/p/distcc
GNU General Public License v2.0
0 stars 0 forks source link

Include server internal error: '<class 'sre_constants.error'>: ('bogus escape (end of line)',)' #35

Closed GoogleCodeExporter closed 9 years ago

GoogleCodeExporter commented 9 years ago
I have a bizarre failure from the include_server, and I'm not sure what I 
should do:

WARNING include server: Preprocessing locally. Include server internal error: 
'<class 
'sre_constants.error'>: ('bogus escape (end of line)',)' for translation unit 
'../source/lib/libport/file-library.cc' ------- Include server stack trace 
-----------

My set up is very unusual: I'm using msvc++ to compile, but I'm using a 
home-grown wrapper 
that takes a gcc-like command line and under the hood actually uses cl.exe.  It 
is sufficiently 
gcc-like to have ccache work perfectly, to have libtool accept it, and to have 
Automake have 
automatic dependency tracking with gcc-like -MT options.

But I have this warning that I don't understand, and I don't know how to track 
it.  I have included 
a portion of the log file that i have when I run

INCLUDE_SERVER_ARGS='-d15' DISTCC_VERBOSE=1
DISTCC_HOSTS='192.168.0.103,lzo,cpp/1' pump make CXX='distcc cl.exe'

Note that the wrapper tries to produce the -v output in order to help the 
include_server find the 
-I path:

build@bf-linux-6   /tmp/./cl.exe foo.cc -v -c
#include "..." search starts here:
#include <...> search starts here:
 /home/build/.wine/dosdevices/c:/vcxx8/VC/include
 /home/build/.wine/dosdevices/c:/vcxx8/VC/PlatformSDK/Include
End of search list.

Distcc also fails on a small-scale experiment here:

DISTCC_VERBOSE=1 DISTCC_HOSTS='192.168.0.103,lzo,cpp/1' pump distcc cl.exe 
foo.cc -c -o 
foo.o |& tee log

(the result of which is the "foo.log" attachement.)

This is with stock 3.1, not the head of svn.
¡

Original issue reported on code.google.com by akim.demaille@gmail.com on 23 Jan 2009 at 3:48

Attachments:

GoogleCodeExporter commented 9 years ago
Thanks for the bug report!

This is an escaping problem in the call to re.sub() from
macro_eval._SubstituteSymbolInString.  It should be escaping backslash escape
sequences in the variable y.

I think the attached patch should fix it.

Original comment by fergus.h...@gmail.com on 23 Jan 2009 at 6:58

Attachments:

GoogleCodeExporter commented 9 years ago

Original comment by fergus.h...@gmail.com on 23 Jan 2009 at 6:58

GoogleCodeExporter commented 9 years ago

Original comment by fergus.h...@gmail.com on 23 Jan 2009 at 7:31

GoogleCodeExporter commented 9 years ago
Fixed in svn revision 657.

Original comment by fergus.h...@gmail.com on 31 Jan 2009 at 9:51