Bug 12623 – gcc-2.3.1-r6 breaks licq-1.2.0a-r2 compilation

• Product: Gentoo Linux
• Status: RESOLVED FIXED

[Description nichao 2002-12-23 11:54:49 UTC]

On Gentoo unstable (ACCEPT_KEYWORDS="~x86"), gcc-2.3.1-r6 breaks licq-1.2.0a-r2
compilation. It works fine with revision -r5 (same CFLAGS), though!

g++ -DHAVE_CONFIG_H -I. -I. -I.. -Wall -I../include -DINSTALL_PREFIX=\"/usr\"  
-fPIC -DPIC -D_REENTRANT  -s -march=athlon-4 -mcpu=athlon-4 -m3dnow -mmmx -msse
-O2 -pipe -fforce-addr -frerun-cse-after-loop -frerun-loop-opt
-falign-functions=4 -mfpmath=sse -mpreferred-stack-boundary=4 -fno-exceptions
-fno-check-new -c icqpacket.cpp
licq.cpp: In member function `int CLicq::Main()':
licq.cpp:684: invalid conversion from `short unsigned int' to `time_t*'
make[2]: *** [licq.o] Error 1
make[2]: *** Waiting for unfinished jobs....
make[2]: Leaving directory `/var/tmp/portage/licq-1.2.0a-r2/work/licq-1.2.0a/src'

I've tried licq's CVS, too, which gives the exact same error.

The gcc changelog says: "Update with patches from Redhat". This is particularly
controversial seeing that some time ago, RedHat patched gcc 2.95.3 and called it
2.96. This patched gcc is known to break stuff - mplayer, for instance.

[Comment 1 Matthew Kennedy (RETIRED) 2002-12-25 16:32:37 UTC]

its not the compiler :) its a type-cast problem in the code which i just
corrected. i'll commit the patch later today 

[Comment 2 Matthew Kennedy (RETIRED) 2002-12-26 02:32:06 UTC]

commited fix in -r3, but marked it ~x86 (unstable/testing). please test and let
me know if problems encountered. if it works for you, reopen and we'll mark it
stable.

[Comment 3 nichao 2002-12-26 13:35:33 UTC]

Thanks - works just fine for me now :-)

[Comment 4 nichao 2002-12-26 13:40:26 UTC]

Reopened as requested. (Didn't know I needed to comment when changing status,
hence this extra comment.)

[Comment 5 Matthew Kennedy (RETIRED) 2002-12-26 22:16:50 UTC]

thanks. marked as stable x86. should be able to rsync the updated -r3 shortly.

Matt