This is the mail archive of the
mailing list for the GCC project.
Re: [PATCH] decide edge's hotness when there is profile info
- From: Dehao Chen <dehao at google dot com>
- To: Xinliang David Li <davidxl at google dot com>
- Cc: Jan Hubicka <hubicka at ucw dot cz>, GCC Patches <gcc-patches at gcc dot gnu dot org>
- Date: Mon, 14 Oct 2013 14:34:04 -0700
- Subject: Re: [PATCH] decide edge's hotness when there is profile info
- Authentication-results: sourceware.org; auth=none
- References: <CAO2gOZVX2QuoUGRy6uRiT4A+=xdeCXmq04tF0Qr1kqezP-y5-A at mail dot gmail dot com> <CAAkRFZLHHDgtfCQQE06O8NKMfSzQuXyweHSoBUwTWseUu7KTnQ at mail dot gmail dot com> <CAO2gOZXwT1Nu4z5NKr1y6xkY2o1dQ6rNxvhT5ap1Kv-WAEo=Pg at mail dot gmail dot com> <20131014194908 dot GA16717 at kam dot mff dot cuni dot cz> <CAO2gOZU8rdLYJm240xz1T=s_+uOx=DV4nM8YEzBHMY9rC5dw4A at mail dot gmail dot com> <CAAkRFZJWSXd6wkVXwsvPf_zkjY=5DF0YhRL6LCnkPn1qttebUw at mail dot gmail dot com>
For my test case, the entire inline instance is optimized away, so
there is no info about it in the profile. I can do some fixup in the
On Mon, Oct 14, 2013 at 2:27 PM, Xinliang David Li <email@example.com> wrote:
> Is it possible to update the callee node summary after profile
> annotate (using information from inline instances which are not
> inlined in early inline)?
> On Mon, Oct 14, 2013 at 2:18 PM, Dehao Chen <firstname.lastname@example.org> wrote:
>> On Mon, Oct 14, 2013 at 12:49 PM, Jan Hubicka <email@example.com> wrote:
>>>> Not for instrumented FDO (not as I know of). But for AutoFDO, this
>>>> could be a potential risk because some callee is marked unlikely
>>>> executed simply because they are inlined and eliminated in the O2
>>>> binary. But in ipa-inline it will not get inlined because the edge is
>>>> not hot from cgraph_maybe_hot_edge_p (because callee is
>>>> UNLIKELY_EXECUTED), while the edge->count is actually hot.
>>> Can't you prevent setting calle to UNLIKELY_EXECUTED in these cases instead?
>>> It seems that having profile set incorrectly will lead to other problems later, too.
>>> We discussed similar problem with Teresa about the missing profiles for comdat,
>>> basically one should detect these cases as profile being lost and go with guessed
>>> profile. (I believe patch for that was posted, too, and so far it seems best approach
>>> to this issue)
>> The current AutoFDO implementation will take all functions that do not
>> have have profile as normally executed, thus use guessed profile for
>> it. This is like using profile for truly hot functions, and using O2
>> for other functions. This works fine. However, it leads to larger code
>> size (approximately 10%~20% larger than FDO).
>> I'd like to introduce another mode for users who care about both
>> performance and code size, and can be sure that profile is
>> representative. In this mode, we will mark all functions without
>> sample as "unlikely executed". However, because AutoFDO use debug info
>> (of optimized code) to represent profile, it's possible that some hot
>> functions (say foo) are inlined and fully eliminated into another hot
>> function (say bar). So in the profile, bar is cold, and because the
>> profile for foo::bar is eliminated, bar will not be inlined into foo
>> before the profile annotation. However, after profile annotate, we can
>> infer from the bb count that foo->bar is hot, thus it should be
>> inlined in ipa-inline phase. However, because bar itself is marked
>> UNLIKELY_EXECUTED, it will not be inlined.
>> One possible workaround would be that during rebuild_cgraph_edges, if
>> we find an edge's callee is unlikely executed, add the edge count to
>> the callee's count and recalculate callee's frequency.