Bug 180304 – gcc-4.2.0 uses a ton of memory while compiling some apps

• Product: Gentoo Linux
• Status: RESOLVED FIXED

[Description Bugs Bunny 2007-05-30 01:17:26 UTC]

There is a major regression in gcc-4.2.0, which upstream has admitted to. They have put up a patch that I have updated with the modifications needed to prevent the ice when building fortran.


Reproducible: Always

Steps to Reproduce:
1. update gcc-4.2.0
2. emerge xorg-server
3.

Actual Results:  
xorg-server uses all avaliable memory and swap, can take up to 8 hours to build on a decent system.

Expected Results:  
regression should be fixed as it prevents testing.

[Comment 1 Bugs Bunny 2007-05-30 01:18:53 UTC]

Created attachment 120667 [details]
gcc-4.2.0-x86_64-regression.patch

All info on upstream bug.

http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30052

[Comment 2 SpanKY 2007-05-30 07:29:30 UTC]

saying "a regression" is not trackable by any means

[Comment 3 SpanKY 2007-06-07 01:18:01 UTC]

*** Bug 181157 has been marked as a duplicate of this bug. ***

[Comment 4 Jakub Moc (RETIRED) 2007-06-14 11:42:26 UTC]

*** Bug 182015 has been marked as a duplicate of this bug. ***

[Comment 5 SpanKY 2007-06-18 03:18:05 UTC]

added the change that actually went into the branch:
http://gcc.gnu.org/viewcvs?view=rev&revision=125227

[Comment 6 Bugs Bunny 2007-06-19 02:05:10 UTC]

(In reply to comment #5)
> added the change that actually went into the branch:
> http://gcc.gnu.org/viewcvs?view=rev&revision=125227
> 

It would be appreciated if a changelog was made that documents all changes instead of making users guess as to what has changed.

[Comment 7 SpanKY 2007-06-19 06:53:47 UTC]

why dont you read the changelog in the patchset

[Comment 8 Jakub Moc (RETIRED) 2007-09-06 12:00:36 UTC]

*** Bug 191487 has been marked as a duplicate of this bug. ***

[Comment 9 Jakub Moc (RETIRED) 2007-09-20 15:27:09 UTC]

*** Bug 192799 has been marked as a duplicate of this bug. ***

[Comment 10 Jakub Moc (RETIRED) 2007-09-20 16:10:59 UTC]

(In reply to comment #7)
> why dont you read the changelog in the patchset

Why don't you just revbump the thing as you are supposed to do in the first place - instead of letting users hit this bug over and over again because you've just silently changed the ebuild to use a new patchset?
 

[Comment 11 SpanKY 2007-09-20 16:24:53 UTC]

i dont revbump things that were not keyworded ... users who want to play with things that arent meant to be played with get to track things themselves