Ping?
https://gcc.gnu.org/ml/gcc-patches/2016-08/msg00872.html
Thanks,
Kugan
On 11 August 2016 at 09:09, kugan wrote:
> Hi,
>
>
> On 10/08/16 20:28, Richard Biener wrote:
>>
>> On Wed, Aug 10, 2016 at 10:57 AM, Jakub Jelinek wrote:
>>>
>>> On Wed, Aug
Hi,
On 19 August 2016 at 21:41, Richard Biener wrote:
> On Tue, Aug 16, 2016 at 9:45 AM, kugan
> wrote:
>> Hi Richard,
>>
>> On 12/08/16 20:43, Richard Biener wrote:
>>>
>>> On Wed, Aug 3, 2016 at 3:17 AM, kugan
>>> wrote:
>>
>>
&g
ant here. The analysis phase should
> not determine
> anything if function is reachable non-locally.
Removed it.
>> +/* Info about value ranges. */
>> +
>> +struct GTY(()) ipa_vr
>> +{
>> + /* The data fields below are valid only if known is true. */
>
Hi Richard,
On 25 August 2016 at 22:24, Richard Biener wrote:
> On Thu, Aug 11, 2016 at 1:09 AM, kugan
> wrote:
>> Hi,
>>
>>
>> On 10/08/16 20:28, Richard Biener wrote:
>>>
>>> On Wed, Aug 10, 2016 at 10:57 AM, Jakub Jelinek wrote:
>>&
Ping ?
Thanks,
Kugan
On 23 August 2016 at 12:11, Kugan Vivekanandarajah
wrote:
> Hi,
>
> On 19 August 2016 at 21:41, Richard Biener wrote:
>> On Tue, Aug 16, 2016 at 9:45 AM, kugan
>> wrote:
>>> Hi Richard,
>>>
>>> On 12/08/16 20:43, Richard Bie
follow some consistent usage here.
It might be also good to gave a FOR_EACH_SSAVAR iterator as we do in
other case. Here is attempt to do this based on what is done in other
places. Bootstrapped and regression tested on X86_64-linux-gnu with no
new regressions. is this OK?
Thanks,
Kugan
gcc
Hi Richard,
On 5 September 2016 at 17:57, Richard Biener wrote:
> On Mon, Sep 5, 2016 at 7:26 AM, Kugan Vivekanandarajah
> wrote:
>> Hi All,
>>
>> While looking at gcc source, I noticed that we are iterating over SSA
>> variable from 0 to num_ssa_names in some case
Hi Richard,
On 6 September 2016 at 19:08, Richard Biener wrote:
> On Tue, Sep 6, 2016 at 2:24 AM, Kugan Vivekanandarajah
> wrote:
>> Hi Richard,
>>
>> On 5 September 2016 at 17:57, Richard Biener
>> wrote:
>>> On Mon, Sep 5, 2016 at 7:26 AM, Kugan V
Hi Richard,
On 6 September 2016 at 19:57, Richard Biener wrote:
> On Tue, Sep 6, 2016 at 11:33 AM, Kugan Vivekanandarajah
> wrote:
>> Hi Richard,
>>
>> On 6 September 2016 at 19:08, Richard Biener
>> wrote:
>>> On Tue, Sep 6, 2016 at 2:24 AM, Kug
Hi Richard,
On 6 September 2016 at 19:08, Richard Biener wrote:
> On Tue, Sep 6, 2016 at 2:24 AM, Kugan Vivekanandarajah
> wrote:
>> Hi Richard,
>>
>> On 5 September 2016 at 17:57, Richard Biener
>> wrote:
>>> On Mon, Sep 5, 2016 at 7:26 AM, Kugan V
insert.
3. In rewrite_expr_tree_parallel, build_and_add_sum relies on either
of operand being inserted. If that is not the case, we have to insert
the stmt_to_insert before calling build_and_add_sum.
4. I also moved all the other stmt_to_insert insertion after the use
stmt are created.
Also regression t
Hi Jakub,
On 26 May 2016 at 18:18, Jakub Jelinek wrote:
> On Thu, May 26, 2016 at 02:17:56PM +1000, Kugan Vivekanandarajah wrote:
>> --- a/gcc/tree-ssa-reassoc.c
>> +++ b/gcc/tree-ssa-reassoc.c
>> @@ -3767,8 +3767,10 @@ swap_ops_for_binary_stmt (vec ops,
>>
trunk if the testing is fine ?
Thanks,
Kugan
gcc/ChangeLog:
2016-05-28 Kugan Vivekanandarajah
* tree-ssa-reassoc.c (swap_ops_for_binary_stmt): Fix swap such that
all fields including stmt_to_insert are swapped.
gcc/testsuite/ChangeLog:
2016-05-28 Kugan Vivekanandarajah
* gcc.dg/tree-ssa
PRs
Thanks,
Kugan
gcc/testsuite/ChangeLog:
2016-05-28 Kugan Vivekanandarajah
* gcc.dg/tree-ssa/pr71269.c: New test.
gcc/ChangeLog:
2016-05-28 Kugan Vivekanandarajah
* tree-ssa-reassoc.c (insert_stmt_before_use): Use find_insert_point so
that inserted stmt will not dominate
Ping ?
I see that Jim has clarified the comments from Andrew.
Thanks,
Kugan
On 13 October 2017 at 08:48, Jim Wilson wrote:
> On Fri, 2017-09-22 at 14:11 -0700, Andrew Pinski wrote:
>> On Fri, Sep 22, 2017 at 11:39 AM, Jim Wilson
>> wrote:
>> >
>> > On Fri,
Hi Jim,
On 1 November 2017 at 03:12, Jim Wilson wrote:
> On Tue, 2017-10-31 at 14:35 +1100, Kugan Vivekanandarajah wrote:
>> Ping ?
>>
>> I see that Jim has clarified the comments from Andrew.
>
> Andrew also suggested that we add a testcase to the testsuite. I
>
Hi,
On 1 November 2017 at 03:12, Jim Wilson wrote:
> On Tue, 2017-10-31 at 14:35 +1100, Kugan Vivekanandarajah wrote:
>> Ping ?
>>
>> I see that Jim has clarified the comments from Andrew.
>
> Andrew also suggested that we add a testcase to the testsuite. I
>
Hi,
Attached patch implements the vld1_*_x2 intrinsics as defined by the
neon document.
Bootstrap for the latest patch is ongoing on aarch64-linux-gnu. Is
this OK for trunk if no regressions?
Thanks,
Kugan
gcc/ChangeLog:
2017-11-06 Kugan Vivekanandarajah
* config/aarch64/aarch64
Ping ?
Thanks,
Kugan
On 27 June 2017 at 11:20, Kugan Vivekanandarajah
wrote:
> https://gcc.gnu.org/ml/gcc-patches/2016-03/msg00614.html added this
> workaround to get kernel building with when TARGET_FIX_ERR_A53_843419
> is enabled.
>
> This was added to support building
Ping^2?
Thanks,
Kugan
On 21 July 2017 at 20:12, Kugan Vivekanandarajah
wrote:
> Ping ?
>
> Thanks,
> Kugan
>
> On 27 June 2017 at 11:20, Kugan Vivekanandarajah
> wrote:
>> https://gcc.gnu.org/ml/gcc-patches/2016-03/msg00614.html added this
>> workaround
ping^3
Thanks,
Kugan
On 11 August 2017 at 16:09, Kugan Vivekanandarajah
wrote:
> Ping^2?
>
> Thanks,
> Kugan
>
> On 21 July 2017 at 20:12, Kugan Vivekanandarajah
> wrote:
>> Ping ?
>>
>> Thanks,
>> Kugan
>>
>> On 27 June 2017 at 11:20,
Hi James,
On 29 August 2017 at 21:31, James Greenhalgh wrote:
> On Tue, Jun 27, 2017 at 11:20:02AM +1000, Kugan Vivekanandarajah wrote:
>> https://gcc.gnu.org/ml/gcc-patches/2016-03/msg00614.html added this
>> workaround to get kernel building with when TARGET_FIX_ERR_A53_84341
g the internal function for this to some extend but for some
cases we should be able to say while in loop distribution itself that
the control flow will not result in loop being vectorized.
Btw, did you run Spec2006 with this? Any notable changes ?
Thanks,
Kugan
On 2 June 2017 at 21:51, Bin Cheng
.
Thanks,
Kugan
This patch adds separate params for rtl unroller so that they can be
tunned accordingly. Default values I have are based on some testing on
aarch64. I am happy to leave it as the current value and set them in
the back-end.
Thanks,
Kugan
gcc/ChangeLog:
2017-09-12 Kugan Vivekanandarajah
This patch adds number of hw prefetchers available to
cpu_prefetch_tune so it can be used in loop unrolling decisions.
Thanks,
Kugan
gcc/ChangeLog:
2017-09-12 Kugan Vivekanandarajah
* config/aarch64/aarch64-protos.h (struct cpu_prefetch_tune): Add
new field hw_prefetchers_avail
This patch prevent tree unroller from completely unrolling inner loops if that
results in excessive strided-loads in outer loop.
Thanks,
Kugan
gcc/ChangeLog:
2017-09-12 Kugan Vivekanandarajah
* config/aarch64/aarch64.c (count_mem_load_streams): New.
(aarch64_ok_to_unroll): New
Change iv_analyze_result to take const_rtx. This is just to make the
next patch compile. No functional changes:
Thanks,
Kugan
gcc/ChangeLog:
2017-09-12 Kugan Vivekanandarajah
* cfgloop.h (iv_analyze_result): Change 2nd param from rtx to
const_rtx.
* df-core.c (df_find_def
This patch adds aarch64_loop_unroll_adjust to limit partial unrolling
in rtl based on strided-loads in loop.
Thanks,
Kugan
gcc/ChangeLog:
2017-09-12 Kugan Vivekanandarajah
* cfgloop.h (iv_analyze_biv): export.
* loop-iv.c: Likewise.
* config/aarch64/aarch64.c (strided_load_p
Hi Andrew,
On 15 September 2017 at 13:20, Andrew Pinski wrote:
> On Thu, Sep 14, 2017 at 6:28 PM, Kugan Vivekanandarajah
> wrote:
>> This patch adds number of hw prefetchers available to
>> cpu_prefetch_tune so it can be used in loop unrolling decisions.
>
> Can yo
Hi Ramana
On 15 September 2017 at 18:40, Ramana Radhakrishnan
wrote:
> On Fri, Sep 15, 2017 at 2:33 AM, Kugan Vivekanandarajah
> wrote:
>> This patch adds aarch64_loop_unroll_adjust to limit partial unrolling
>> in rtl based on strided-loads in loop.
>>
>> Thanks
Hi Andrew,
On 15 September 2017 at 13:36, Andrew Pinski wrote:
> On Thu, Sep 14, 2017 at 6:33 PM, Kugan Vivekanandarajah
> wrote:
>> This patch adds aarch64_loop_unroll_adjust to limit partial unrolling
>> in rtl based on strided-loads in loop.
>
> Can you expand on th
Hi Richard,
On 15 September 2017 at 19:31, Richard Biener
wrote:
> On Fri, Sep 15, 2017 at 3:27 AM, Kugan Vivekanandarajah
> wrote:
>> This patch adds separate params for rtl unroller so that they can be
>> tunned accordingly. Default values I have are based on some testing o
Hi Richard,
On 18 September 2017 at 17:50, Richard Biener
wrote:
> On Mon, Sep 18, 2017 at 3:36 AM, Kugan Vivekanandarajah
> wrote:
>> Hi Richard,
>>
>> On 15 September 2017 at 19:31, Richard Biener
>> wrote:
>>> On Fri, Sep 15, 2017 at 3:27 AM, Kugan
Hi Richard,
On 7 September 2016 at 19:35, Richard Biener wrote:
> On Wed, Sep 7, 2016 at 2:21 AM, Kugan Vivekanandarajah
> wrote:
>> Hi Richard,
>>
>> On 6 September 2016 at 19:08, Richard Biener
>> wrote:
>>> On Tue, Sep 6, 2016 at 2:24 AM, Kugan Vivek
based on the
feedback.
Please let me know what you thing.
Thanks,
Kugan
From 332e0e9f938c6af50e826d8224d07ebf3678a0e0 Mon Sep 17 00:00:00 2001
From: Kugan Vivekanandarajah
Date: Fri, 13 May 2016 13:41:01 +1000
Subject: [PATCH 4/4] Add new type promotion pass
---
gcc/ChangeLog
Hi,
static variable all_extensions in aarch64.c is not used and therefore
dead. I don’t see any reason why it should be there. Attached patch
removes this.
Bootstrapped on aarch64-linux-gnu. Regression testing is ongoing.
Is this OK for trunk?
Thanks,
Kugan
gcc/ChangeLog:
2016-05-17 Kugan
+ || real_minus_onep (last->op))
Is this Still OK. Bootstrap and regression testing on ARM, AARCH64 and
x86-64 didn’t have any new regressions.
Thanks,
Kugan
diff --git a/gcc/testsuite/gcc.dg/tree-ssa/pr40921.c
b/gcc/testsuite/gcc.dg/tree-ssa/pr40921.c
index e69de29..3a5a23a 100644
--- a
We could try Martin Liška's approach, We could also move _17 = c_7(D)
* 3; at tree-ssa-reassoc.c:3897 satisfy the gcc_assert. We could do
this based on the use count of _17.
This patch does this. I have no preferences. Any thoughts ?
Thanks,
Kugan
On 19 May 2016 at 18:04, Martin Lišk
On 19 May 2016 at 18:55, Richard Biener wrote:
> On Thu, May 19, 2016 at 10:26 AM, Kugan
> wrote:
>> Hi,
>>
>>
>> On 19/05/16 18:21, Richard Biener wrote:
>>> On Thu, May 19, 2016 at 10:12 AM, Kugan Vivekanandarajah
>>> wrote:
>>>> Hi
erting the multiplication to
> rewrite_expr_tree time. For example by adding a ops->stmt_to_insert
> member.
>
Here is an implementation based on above. Bootstrap on x86-linux-gnu
is OK. regression testing is ongoing.
Thanks,
Kugan
gcc/ChangeLog:
2016-05-20 Kugan Vivekanandarajah
tested on x86-64-linux-gnu with no new regressions.
Is this OK for trunk?
Thanks,
Kugan
gcc/testsuite/ChangeLog:
2016-05-20 Kugan Vivekanandarajah
* gcc.dg/tree-ssa/pr71179.c: New test.
gcc/ChangeLog:
2016-05-20 Kugan Vivekanandarajah
* tree-ssa-reassoc.c
On 20 May 2016 at 21:07, Richard Biener wrote:
> On Fri, May 20, 2016 at 1:51 AM, Kugan Vivekanandarajah
> wrote:
>> Hi Richard,
>>
>>> I think it should have the same rank as op or op + 1 which is the current
>>> behavior. Sth else doesn't work c
Hi Jeff,
On 20 May 2016 at 04:17, Jeff Law wrote:
> On 05/15/2016 06:45 PM, Kugan Vivekanandarajah wrote:
>>
>> Hi Richard,
>>
>> Now that stage1 is open, I would like to get the type promotion passes
>> reviewed again. I have tested the patches on aarch64, x86-6
(optimized).
I will also try to gather test-cases based on testing/benchmarking.
Thanks,
Kugan
On 23 May 2016 at 21:35, Richard Biener wrote:
> On Sat, May 21, 2016 at 8:08 AM, Kugan Vivekanandarajah
> wrote:
>> On 20 May 2016 at 21:07, Richard Biener wrote:
>>> On Fri, May 20, 2016 at 1:51 AM, Kugan Vivekanandarajah
>>> wrote:
>>>> Hi Richard,
On 24 May 2016 at 18:36, Christophe Lyon wrote:
> On 24 May 2016 at 05:13, Kugan Vivekanandarajah
> wrote:
>> On 23 May 2016 at 21:35, Richard Biener wrote:
>>> On Sat, May 21, 2016 at 8:08 AM, Kugan Vivekanandarajah
>>> wrote:
>>>> On 20 May 2016 at
reducing the test-case is
appreciated.
Regression testing on x86_64-linux-gnu and bootstrap didn’t find any new issues.
Is this OK for trunk?
Thanks,
Kugan
gcc/testsuite/ChangeLog:
2016-05-24 Kugan Vivekanandarajah
* gfortran.dg/pr71252.f90: New test.
gcc/ChangeLog:
2016-05-24 Kugan
Hi,
This patch Fixes absfloat16.c testcase to have the dg-add-options float16 at
the correct order. Due to this mixup, this test is failing for some arm
variants.
Is this OK for trunk?
Thanks,
Kugan
0001-Fix-absfloat16.c-testcase.patch
Description: 0001-Fix-absfloat16.c-testcase.patch
ping?
Thanks,
Kugan
From: Kugan Vivekanandarajah
Sent: Tuesday, 20 August 2024 6:18 PM
To: Jakub Jelinek
Cc: gcc-patches@gcc.gnu.org ;
richard.guent...@gmail.com ;
richard.sandif...@arm.com
Subject: Re: [PR middle-end/114635] Set OMP safelen handling to
Hi Richard,
Thanks for the review.
> On 8 Oct 2024, at 7:15 pm, Richard Biener wrote:
>
> External email: Use caution opening links or attachments
>
>
> On Mon, Aug 5, 2024 at 7:05 AM Kugan Vivekanandarajah
> wrote:
>>
>>
>>
>>>
Hi Richard,
Thanks for the review.
> On 25 Oct 2024, at 8:53 pm, Richard Biener wrote:
>
> External email: Use caution opening links or attachments
>
>
> On Fri, Oct 25, 2024 at 12:22 AM Kugan Vivekanandarajah
> wrote:
>>
&g
Hi Richard,
Thanks for the review.
> On 28 Oct 2024, at 9:18 pm, Richard Biener wrote:
>
> External email: Use caution opening links or attachments
>
>
> On Mon, Oct 28, 2024 at 9:35 AM Kugan Vivekanandarajah
> wrote:
>>
>> Hi,
>>
>> When ifcvt
> On 31 Oct 2024, at 6:18 pm, Jakub Jelinek wrote:
>
> External email: Use caution opening links or attachments
>
>
> On Tue, Oct 29, 2024 at 05:01:40AM +, Kugan Vivekanandarajah wrote:
>> For param_vect_max_version_for_alias_checks of 15, the average code si
> On 31 Oct 2024, at 7:29 pm, Jakub Jelinek wrote:
>
> External email: Use caution opening links or attachments
>
>
> On Thu, Oct 31, 2024 at 08:21:09AM +, Kugan Vivekanandarajah wrote:
>>
>>
>>> On 31 Oct 2024, at 6:18 pm, Jakub Jelinek wrote:
>
different
from general dont_vectorize) specifically for loops versioned. BB vectorization
does not need to honour this and still can vectorize.
Bootstrapped and regression tested on aarch64-linux-gnu with no new regressions.
Is this OK?
Thanks,
Kugan
0001-PATCH-Fix-SLP-when-ifcvt-versioned-loop-is
f at teast 11 where as the current
default is 10.
Bootstrapped and regression tested on aarc64-linux-gnu with no new regressions.
Thanks,
Kugan
0001-RFC-PATCH-Adjust-param_vect_max_version_for_alias_ch.patch
Description: 0001-RFC-PATCH-Adjust-param_vect_max_version_for_alias_ch.patch
h one insm. Hence,
when the operands are equal, split after reload.
Bootstrapped and recession tested on aarch64-linux-gnu, Is this ok for trunk?
Thanks,
Kugan
0001-PATCH-AARCH64-PR115258-Fix-excess-moves.patch
Description: 0001-PATCH-AARCH64-PR115258-Fix-excess-moves.patch
hen I
force the loop to unroll for x86. Thus, to keep it simple, moving the test to
gcc.target/aarch64.
Regression tested on aarch64-linux-gnu. Is this OK?
Thanks,
Kugan
0001-testsuite-Fix-bb-slp-77.c.patch
Description: 0001-testsuite-Fix-bb-slp-77.c.patch
Hi Richard,
> On 29 Oct 2024, at 8:33 pm, Richard Biener wrote:
>
> External email: Use caution opening links or attachments
>
>
> On Tue, Oct 29, 2024 at 9:24 AM Kugan Vivekanandarajah
> wrote:
>>
>> Hi Richard,
>> Thanks for the review.
>>
&
Hi Richard,
> On 17 Sep 2024, at 7:36 pm, Richard Biener wrote:
>
> External email: Use caution opening links or attachments
>
>
> On Tue, Sep 17, 2024 at 10:31 AM Kugan Vivekanandarajah
> wrote:
>>
>> Hi Richard,
>>
>>> On 10 Sep 2024, at 9:33
Ping?
Thanks,
Kugan
> On 2 Nov 2024, at 7:49 pm, Kugan Vivekanandarajah
> wrote:
>
> External email: Use caution opening links or attachments
>
>
> > On 31 Oct 2024, at 7:29 pm, Jakub Jelinek wrote:
> >
> > External email: Use caution opening links or a
Hi Richard,
I want to follow up on this and see if you have a fix for this.
Thanks,
Kugan
> On 29 Oct 2024, at 9:41 pm, Richard Sandiford
> wrote:
>
> External email: Use caution opening links or attachments
>
>
> Kugan Vivekanandarajah writes:
>> Hi,
>
annotate profile for GIMPLE_CALL stmt
and extract BB counts from edge counts.
Regression tested on aarch64-linux-gnu with no new regression.
Also successfully done autoprofiledbootstrap with the relevant patch.
Is this OK for trunk?
Thanks,
Kugan
0001-AUTOFDO-Fix-annotated-profile-for-de
.
Regression tested on aarch64-linux-gnu with no new regression.
Also successfully done autoprofiledbootstrap with the relevant patch.
Is this OK for trunk?
Thanks,
Kugan
0002-AUTOFDO-Merge-profiles-of-clones-before-annotating.patch
Description: 0002-AUTOFDO-Merge-profiles-of-clones-before
new regression.
Also successfully done autoprofiledbootstrap with the relevant patch.
Is this OK for trunk?
Thanks,
Kugan
0004-AUTOFDO-AARCH64-Add-support-for-profilebootstrap.patch
Description: 0004-AUTOFDO-AARCH64-Add-support-for-profilebootstrap.patch
Adding Eugene and Andi to CC as Sam suggested.
> On 13 May 2025, at 12:57 am, Richard Sandiford
> wrote:
>
> External email: Use caution opening links or attachments
>
>
> Kugan Vivekanandarajah writes:
>> diff --git a/configure.ac b/configure.ac
>> inde
Thanks Richard for the review.
> On 20 May 2025, at 2:47 am, Richard Sandiford
> wrote:
>
> External email: Use caution opening links or attachments
>
>
> Kugan Vivekanandarajah writes:
>> diff --git a/Makefile.in b/Makefile.in
>> index b1ed67d3d4f..b5e3e5
> On 16 May 2025, at 12:10 am, Andi Kleen wrote:
>
> External email: Use caution opening links or attachments
>
>
> On Wed, May 14, 2025 at 02:46:15AM +, Kugan Vivekanandarajah wrote:
>> Adding Eugene and Andi to CC as Sam suggested.
>>
>>> On 13 M
> On 26 May 2025, at 2:25 pm, Andrew Pinski wrote:
>
> External email: Use caution opening links or attachments
>
>
> On Tue, May 20, 2025 at 3:09 AM Kugan Vivekanandarajah
> wrote:
>>
>> Thanks Richard for the review.
>>
>>> On 20 May
es and only see if afdo
annotations are there.
Any thoughts?
Thanks,
Kugan
>
> Honza
> <0002-AUTOFDO-Merge-profiles-of-clones-before-annotating.patch>
10:18.479228 1692721 symbol_map.cc:477] Adding loadable exec
> segment: offset=1000 vaddr=401000
>
> Did someone run SPEC recently? I made auto-FDO spec config and tested
> -Ofast with ipa-icf, ipa-cp-clone and ipa-sra disabled (to get rid of
> the clone merging). I get sort of com
> On 26 May 2025, at 2:47 pm, Kugan Vivekanandarajah
> wrote:
>
> External email: Use caution opening links or attachments
>
>
> > On 26 May 2025, at 2:25 pm, Andrew Pinski wrote:
> >
> > External email: Use caution opening links or attachments
>
Ping?
Thanks,
Kugan
> On 9 May 2025, at 11:54 am, Kugan Vivekanandarajah
> wrote:
>
> External email: Use caution opening links or attachments
>
>
> This patch add support for merging profiles from multiple clones.
> That is, when optimized binaries have clones suc
Ping?
Thanks,
Kugan
> On 9 May 2025, at 11:51 am, Kugan Vivekanandarajah
> wrote:
>
> External email: Use caution opening links or attachments
>
>
> This patch fixes wrong annotation of profiles when call statement is
> de-duplicated. i.e., when we may have same st
Ping?
Thanks,
Kugan
> On 9 May 2025, at 11:55 am, Kugan Vivekanandarajah
> wrote:
>
> ipa-split is not now run for auto-profile. IMO this was an oversight.
> This patch enables it similar to PGO runs.
>
> gcc/ChangeLog:
>
>* ipa-split.cc pass_feedback_spl
:
* auto-profile.cc (autofdo_source_profile::read): Dump message
while merging profile.
* pass_manager.h (get_pass_auto_profile): New.
gcc/testsuite/ChangeLog:
* gcc.dg/tree-prof/clone-merge-1.c: New test.
Is this OK?
Thanks,
Kugan
0001-AutoFDO_v2-Profile-merging-for-clone
: Enable autofdo tests for aarch64.
Is this OK?
Thanks,
Kugan
0001-AUTOFDO-Enable-autofdo-tests-for-aarch64.patch
Description: 0001-AUTOFDO-Enable-autofdo-tests-for-aarch64.patch
Hi Honza,
> On 6 Jun 2025, at 6:34 pm, Jan Hubicka wrote:
>
> External email: Use caution opening links or attachments
>
>
>> Kugan Vivekanandarajah writes:
>>> Add support for autoprofiledbootstrap in aarch64.
>>> This is similar to what is done for
one used for gcc/config/).
>
> It is incovenient that the toplevel doesn't have access to the logic
> used to set that variable though...
I changed it to:
+# Special case cpu_type for x86_64 as it shares AUTO_PROFILE from i386.
+if test "${cpu_type}" = "x86_64" ; then
+ cpu_type="i386"
+fs
Is this ok? Tested on x86_64 and aarch64 linux-gnu.
Thanks,
Kugan
>
> Richard
0001-AutoFDO-Fix-profile-bootstrap-for-x86_64.patch
Description: 0001-AutoFDO-Fix-profile-bootstrap-for-x86_64.patch
> On 9 Jun 2025, at 9:43 am, Kugan Vivekanandarajah
> wrote:
>
>
>
> > On 7 Jun 2025, at 3:30 pm, Kugan Vivekanandarajah
> > wrote:
> >
> > Hi,
> >
> >
> > > On 6 Jun 2025, at 4:15 pm, Kugan Vivekanandarajah
> > > wrote
n sticking to the one
> that gcc/configure* already uses (i.e. the one used for gcc/config/).
>
> It is incovenient that the toplevel doesn't have access to the logic
> used to set that variable though...
>
I changed it to:
+# Special case cpu_type for x86_64 as it shares AU
elf. Private (static) functions
with the same name also will have the same issue.
Dhruv is working on an RFC for this.
Thanks,
Kugan
>
> Overwritting the data by the last clone is definitely bad, so the patch
> is OK, but we should figure out what happens in the cases above.
>
>
roll "Peeled loop 2, 1 times”
I also noticed that some tests are only enabled for x86. I am also seeing:
./gcc/testsuite/gcc/gcc.sum:UNSUPPORTED: gcc.dg/tree-prof/pr66295.c
./gcc/testsuite/gcc/gcc.sum:UNSUPPORTED: gcc.dg/tree-prof/split-1.c
./gcc/testsuite/gcc/gcc.sum:UNSUPPORTED: gcc.dg/tree-prof/val-prof-10.c
./gcc/testsuite/gcc/gcc.sum:UNSUPPORTED: gcc.dg/tree-prof/val-prof-7.c
./gcc/testsuite/gcc/gcc.sum:UNSUPPORTED: gcc.dg/tree-prof/pr66295.c
./gcc/testsuite/gcc/gcc.sum:UNSUPPORTED: gcc.dg/tree-prof/split-1.c
./gcc/testsuite/gcc/gcc.sum:UNSUPPORTED: gcc.dg/tree-prof/val-prof-10.c
./gcc/testsuite/gcc/gcc.sum:UNSUPPORTED: gcc.dg/tree-prof/val-prof-7.c
Thanks,
Kugan
> Honza
zero, preserve
quality info. */
- else if (count->nonzero_p ()
+ else if (!count->nonzero_p ()
+ || count->quality () == GUESSED_LOCAL
|| count->quality () == GUESSED)
*count = profile_count::zero ().afdo ();
}
Thanks,
Kugan
>
> Honza
>>
>> Thanks,
>> Kugan
Hi Andrew,
> On 6 Jun 2025, at 8:18 am, Andrew Pinski wrote:
>
> External email: Use caution opening links or attachments
>
>
> On Wed, Jun 4, 2025 at 12:02 AM Kugan Vivekanandarajah
> wrote:
>>
>> This patch introduces a new testcase to verify the mergin
ay?
>>
>> Splitting out inlining as its own phase also means that it can
>> eventually be handed off to ipa-inline to handle, thus making
>> auto-profile independent of early inline. This will simplify the code a
>> fair bit. Is this a good direction to go in?
>
> On 17 Jun 2025, at 4:18 pm, Dhruv Chawla wrote:
>
> On 17/06/25 06:10, Kugan Vivekanandarajah wrote:
>> External email: Use caution opening links or attachments
>> Hi,
>> As discusses earlier, get_original_name is used to match profile binary
>> names to
>
running autoprofiledbootstrap and tree-prof check that
exercises auto-profile pass.
gcc/ChangeLog:
* auto-profile.cc (isAsciiDigit): New.
(get_original_name): Strip suffixes only for compiler generated
names tat happens after auto-profile.
Thanks,
Kugan
0001-AutoFDO-Fix
tialized. */
> struct cgraph_edge *new_edge
> - = indirect_edge->make_speculative (direct_call,
> -profile_count::uninitialized ());
> + = indirect_edge->make_speculative
> + (direct_call,
> +
Hi,
> On 17 Jun 2025, at 4:51 pm, Kugan Vivekanandarajah
> wrote:
>
> External email: Use caution opening links or attachments
>
>
>> On 17 Jun 2025, at 4:18 pm, Dhruv Chawla wrote:
>>
>> On 17/06/25 06:10, Kugan Vivekanandarajah wrote:
>>>
> Given that this is tail-recursive, I feel like recursion is not necessary here
> and it would be more efficient to have this be a loop instead. The
> implementation looks okay as is, though.
IMO doing this in a loop would have to handle all the above cases and
Number of samples to get to the desrired percentile.
Should we also track the branch probability in GCOV. This should be easy to
calculate from perf profille. This may help disambiguate profile counts.
Thanks,
Kugan
>
> seems like useful info to handle autoFDO 0s more orrectly, so
501 - 593 of 593 matches
Mail list logo
