From: Tom de Vries <Tom_deVries@mentor.com>
To: Thomas Schwinge <thomas@codesourcery.com>
Cc: GCC Patches <gcc-patches@gcc.gnu.org>
Subject: Re: [PING][PATCH][PR65443] Add transform_to_exit_first_loop_alt
Date: Mon, 08 Jun 2015 22:12:00 -0000 [thread overview]
Message-ID: <557611B6.3080809@mentor.com> (raw)
In-Reply-To: <87twuigpo8.fsf@kepler.schwinge.homeip.net>
On 08/06/15 17:55, Thomas Schwinge wrote:
> Hi Tom!
>
> On Mon, 8 Jun 2015 12:43:01 +0200, Tom de Vries <Tom_deVries@mentor.com> wrote:
>> There are two problems in try_transform_to_exit_first_loop_alt:
>> 1. In case the latch is not a singleton bb, the function should return
>> false rather than true.
>> 2. The check for singleton bb should ignore debug-insns.
>>
>> Attached patch fixes these problems.
>
>> Fix try_transform_to_exit_first_loop_alt
>
>> PR tree-optimization/66442
>> * gimple-iterator.h (gimple_seq_nondebug_singleton_p): Add function.
>> * tree-parloops.c (try_transform_to_exit_first_loop_alt): Return false
>> if the loop latch is not a singleton. Use
>> gimple_seq_nondebug_singleton_p instead of gimple_seq_singleton_p.
>
> Per my testing, the backport of this patch that you committed to
> gomp-4_0-branch, r224219, introduces a number of regressions in your
> OpenACC kernels test cases, specifically the »scan-tree-dump-times
> parloops_oacc_kernels "(?n)pragma omp target
> oacc_parallel.*num_gangs\\(32\\)" 1« tests. Would you please have a
> look?
>
>
Hi Thomas,
I seem to have committed (to both trunk and gomp-4_0-branch) an older
version of the patch, which contained an incorrect version of
gimple_seq_nondebug_singleton_p.
I'll correct the mistake tomorrow morning.
Thanks,
- Tom
> GrüÃe,
> Thomas
>
>
>> gcc/gimple-iterator.h | 29 +++++++++++++++++++++++++++++
>> gcc/tree-parloops.c | 4 ++--
>> 2 files changed, 31 insertions(+), 2 deletions(-)
>>
>> diff --git a/gcc/gimple-iterator.h b/gcc/gimple-iterator.h
>> index 87e943a..76fa456 100644
>> --- a/gcc/gimple-iterator.h
>> +++ b/gcc/gimple-iterator.h
>> @@ -345,4 +345,33 @@ gsi_seq (gimple_stmt_iterator i)
>> return *i.seq;
>> }
>>
>> +/* Determine whether SEQ is a nondebug singleton. */
>> +
>> +static inline bool
>> +gimple_seq_nondebug_singleton_p (gimple_seq seq)
>> +{
>> + gimple_stmt_iterator gsi;
>> +
>> + /* Find a nondebug gimple. */
>> + gsi.ptr = gimple_seq_first (seq);
>> + gsi.seq = &seq;
>> + gsi.bb = NULL;
>> + while (!gsi_end_p (gsi)
>> + && is_gimple_debug (gsi_stmt (gsi)))
>> + gsi_next (&gsi);
>> +
>> + /* No nondebug gimple found, not a singleton. */
>> + if (gsi_end_p (gsi))
>> + return false;
>> +
>> + /* Find a next nondebug gimple. */
>> + gsi_next (&gsi);
>> + while (!gsi_end_p (gsi)
>> + && is_gimple_debug (gsi_stmt (gsi)))
>> + gsi_next (&gsi);
>> +
>> + /* Only a singleton if there's no next nondebug gimple. */
>> + return gsi_end_p (gsi);
>> +}
>> +
>> #endif /* GCC_GIMPLE_ITERATOR_H */
>> diff --git a/gcc/tree-parloops.c b/gcc/tree-parloops.c
>> index 02f44eb..c4b83fe 100644
>> --- a/gcc/tree-parloops.c
>> +++ b/gcc/tree-parloops.c
>> @@ -1769,8 +1769,8 @@ try_transform_to_exit_first_loop_alt (struct loop *loop,
>> tree nit)
>> {
>> /* Check whether the latch contains a single statement. */
>> - if (!gimple_seq_singleton_p (bb_seq (loop->latch)))
>> - return true;
>> + if (!gimple_seq_nondebug_singleton_p (bb_seq (loop->latch)))
>> + return false;
>>
>> /* Check whether the latch contains the loop iv increment. */
>> edge back = single_succ_edge (loop->latch);
>> --
>> 1.9.1
>>
next prev parent reply other threads:[~2015-06-08 22:06 UTC|newest]
Thread overview: 13+ messages / expand[flat|nested] mbox.gz Atom feed top
2015-03-27 14:10 [PATCH, stage1][PR65443] " Tom de Vries
2015-04-03 12:40 ` Tom de Vries
2015-04-15 13:39 ` [PING][PATCH][PR65443] " Tom de Vries
2015-04-20 12:26 ` Richard Biener
2015-05-14 11:47 ` Tom de Vries
2015-05-26 11:05 ` Richard Biener
2015-06-04 8:37 ` Tom de Vries
2015-06-08 10:44 ` Tom de Vries
2015-06-08 11:08 ` Richard Biener
2015-06-08 15:59 ` Thomas Schwinge
2015-06-08 22:12 ` Tom de Vries [this message]
2015-06-09 7:47 ` Tom de Vries
2015-06-01 8:12 ` [gomp4,committed,PR65443] Add transform_to_exit_first_loop_al Tom de Vries
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=557611B6.3080809@mentor.com \
--to=tom_devries@mentor.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=thomas@codesourcery.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).