make fails with editline errors

[GoogleCodeExporter]

What steps will reproduce the problem?
1. ./configure --prefix=/usr
2. make

What is the expected output? What do you see instead?

make should succeed, but instead it fails:

 gcc -DHAVE_CONFIG_H -I. -I../.. -I../../../../third_party/editline
-D__STDC_CONSTANT_MACROS -D_DEBUG
-DGGL_MODULE_DIR=\"/usr/local/lib/google-gadgets/modules\"
-DGGL_INCLUDE_DIR=\"/usr/local/include/google-gadgets\"
-DGGL_SYSDEPS_INCLUDE_DIR=\"/usr/local/lib/google-gadgets/include\"
-DGGL_LIBEXEC_DIR=\"/usr/local/lib/google-gadgets\"
-DGGL_RESOURCE_DIR=\"/usr/local/share/google-gadgets\" -DGGL_HOST_LINUX=1
-DHAVE_X11=1 -DMOZILLA_FIVE_HOME=\"/usr/lib/xulrunner-1.8.1\"
-DHAVE_PTHREAD=1 -DSYS_UNIX -DHAVE_STDLIB -DUNIQUE_HISTORY -DUSE_DIRENT
-DHIDE -DHAVE_TCGETATTR -DANSI_ARROWS -g -O0 -Werror -Wall -Wconversion
-Wno-implicit -Wno-pointer-sign -Wno-conversion -MT
libeditline_la-editline.lo -MD -MP -MF .deps/libeditline_la-editline.Tpo -c
../../../../third_party/editline/editline.c  -fPIC -DPIC -o
.libs/libeditline_la-editline.o
cc1: warnings being treated as errors
../../../../third_party/editline/editline.c: In function 'TTYflush':
../../../../third_party/editline/editline.c:173: error: call to function
'write' without a real prototype
../../../../third_party/editline/editline.c:158: note: 'write' was declared
here
../../../../third_party/editline/editline.c: In function 'TTYget':
../../../../third_party/editline/editline.c:238: error: call to function
'read' without a real prototype
../../../../third_party/editline/editline.c:157: note: 'read' was declared here
../../../../third_party/editline/editline.c: In function 'readline':
../../../../third_party/editline/editline.c:1021: error: call to function
'rl_ttyset' without a real prototype
../../../../third_party/editline/editline.h:118: note: 'rl_ttyset' was
declared here
../../../../third_party/editline/editline.c:1032: error: call to function
'rl_ttyset' without a real prototype
../../../../third_party/editline/editline.h:118: note: 'rl_ttyset' was
declared here
make[3]: *** [libeditline_la-editline.lo] Error 1
make[3]: Leaving directory
`/home/sandy/tarballs/google-gadgets-for-linux-0.9.1/build/debug/third_party/edi
tline'
make[2]: *** [all-recursive] Error 1
make[2]: Leaving directory
`/home/sandy/tarballs/google-gadgets-for-linux-0.9.1/build/debug/third_party'
make[1]: *** [all-recursive] Error 1
make[1]: Leaving directory
`/home/sandy/tarballs/google-gadgets-for-linux-0.9.1/build/debug'
make: *** [all] Error 2

What version of the product are you using? On what operating system?

Using the 0.9.1 tarball on openSUSE 11 RC1.

Please provide any additional information below.

Let me know if you need other info.

Original issue reported on code.google.com by sanforda...@gmail.com on 3 Jun 2008 at 6:24

[GoogleCodeExporter]

I should mention that this particular output is from following the HowToBuild 
wiki
page instructions and building in build/debug, but the errors were the same 
when I
just ran ./configure --prefix=/usr && make from the top-level source directory 
of the
0.9.1 tarball.

Original comment by sanforda...@gmail.com on 3 Jun 2008 at 6:26

[GoogleCodeExporter]

it is because the tarball ship with -Werror by default.

attached is a patch that don't set it for non debug build.

Original comment by hfigui...@gmail.com on 4 Jun 2008 at 9:13

Attachments:

• google-gadgets-configure.patch

[GoogleCodeExporter]

I applied hfiguiere's patch to the tarball and had another build problem 
(something
about an undeclared rand function, I think), so I switched to building from SVN 
and
haven't experienced this problem since.

Original comment by sanforda...@gmail.com on 4 Jun 2008 at 10:03

[GoogleCodeExporter]

see bug 120 for the rest (there is a patch). it is a different issue related to 
the
fact that we use gcc 4.3 on SuSE 11.

Original comment by hfigui...@gmail.com on 4 Jun 2008 at 11:54

[GoogleCodeExporter]

also trunk has a --disable-werror now. I'd rather have the flag the other way, 
as an
opt-in. or a default for non-debug that is no.

Original comment by hfigui...@gmail.com on 5 Jun 2008 at 3:32

[GoogleCodeExporter]

0.9.2 as this, and even if not ideal, it is satisfactory.

I think you can close this issue.

Original comment by hfigui...@gmail.com on 7 Jun 2008 at 2:42

[GoogleCodeExporter]

Thanks for your report.

Original comment by james...@gmail.com on 7 Jun 2008 at 3:03

• Changed state: Fixed