This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: C++ optimization: compile time + memory consumption regressionon gcc3.3 branch
- From: Zack Weinberg <zack at codesourcery dot com>
- To: Karel Gardas <kgardas at objectsecurity dot com>
- Cc: GCC Mailing List <gcc at gcc dot gnu dot org>
- Date: Mon, 03 Mar 2003 10:25:11 -0800
- Subject: Re: C++ optimization: compile time + memory consumption regressionon gcc3.3 branch
- References: <Pine.LNX.4.43.0303031857180.32419-100000@thinkpad.c0202001.roe.itnq.net>
Karel Gardas <kgardas at objectsecurity dot com> writes:
> On Mon, 3 Mar 2003, Zack Weinberg wrote:
>
>> Karel Gardas <kgardas at objectsecurity dot com> writes:
>>
>> > % cumulative self self total
>> > time seconds seconds calls Ks/call Ks/call name
>> > 32.39 370.58 370.58 737172489 0.00 0.00 fixup_var_refs_1
>> > 30.24 716.59 346.01 177958 0.00 0.00 fixup_var_refs_insns
>> > 15.41 892.87 176.28 737172497 0.00 0.00 fixup_var_refs_insn
>> > 4.54 944.81 51.94 242350577 0.00 0.00 reg_mentioned_p
>> > 2.63 974.88 30.07 16240 0.00 0.00 fixup_var_refs
>> > 1.45 991.47 16.59 245604291 0.00 0.00 rtx_equal_p
>>
>> This is a well-known quadratic bottleneck -- unfortunately several
>> bright people have tried and failed to fix it.
>
> Thanks for the info, but it leads me to the question: is it right that it
> was introduced between 3.2 and 3.3 branch points?
The bottleneck itself has been there since forever, but changes in the
3.3 timeframe may have caused it to trigger on your test case where it
did not formerly. I do not know for sure.
> BTW: Since it is ``well-known quadratic bottleneck'' I assume that GNAT
> already contains some proper bugreport, is it ok, or shall I submit one?
I don't know. If you would like to look through GNATS for one and
submit one if there is not already (with preprocessed test case) that
would be helpful.
zw