This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug middle-end/54146] New: Very slow compile at -O1 (expand vars)
- From: "glisse at gcc dot gnu.org" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: Tue, 31 Jul 2012 18:33:30 +0000
- Subject: [Bug middle-end/54146] New: Very slow compile at -O1 (expand vars)
- Auto-submitted: auto-generated
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=54146
Bug #: 54146
Summary: Very slow compile at -O1 (expand vars)
Classification: Unclassified
Product: gcc
Version: 4.8.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: middle-end
AssignedTo: unassigned@gcc.gnu.org
ReportedBy: glisse@gcc.gnu.org
Created attachment 27912
--> http://gcc.gnu.org/bugzilla/attachment.cgi?id=27912
testcase
Hello,
trying to compile the attached program on x86_64, g++ slow.cc -frounding-math
-c -w -O2 fails after about an hour. The memory consumption (as shown by top)
never went above 1.5G. clang compiles the whole thing in 1 minute (well, it
won't compile the preprocessed source attached here, but the original program).
If you remove most of the check<...> lines at then end (keep only the first
one), the program takes about 10s to compile at -O0, and 30s at -O1 (or -O2,
-O3), with 2/3 of the time spent in "expand vars" according to -ftime-report.
The most relevant report I found in bugzilla is PR 38474, but it seems to say
that expand vars should be fast now...