This is the mail archive of the gcc-patches@gcc.gnu.org mailing list for the GCC project.

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]
Other format:	[Raw text]

Re: [RFC] introduce --param max-lto-partition for having an upper bound on partition size

From: Prathamesh Kulkarni <prathamesh dot kulkarni at linaro dot org>
To: Jan Hubicka <hubicka at ucw dot cz>
Cc: Richard Biener <rguenther at suse dot de>, gcc Patches <gcc-patches at gcc dot gnu dot org>, Ramana Radhakrishnan <ramana dot radhakrishnan at arm dot com>
Date: Mon, 4 Apr 2016 18:57:47 +0530
Subject: Re: [RFC] introduce --param max-lto-partition for having an upper bound on partition size
Authentication-results: sourceware.org; auth=none
References: <CAAgBjMkjMcf7XTvBOtnTr-zcKu-rED3WcLY=a4iYoijOw3V3vQ at mail dot gmail dot com> <C74AF7F0-5036-4524-99BE-E24A4243FC88 at suse dot de> <CAAgBjM=gBmgxQCXnvvjXYpVoU5p4wmkpkO0tMepjzz7yiB38vw at mail dot gmail dot com> <alpine dot LSU dot 2 dot 11 dot 1604041021170 dot 13384 at t29 dot fhfr dot qr> <CAAgBjMk=qWCt8VB7f_4+x-Ck6OwsV_CXtnyX1QGKuqta4sPKYA at mail dot gmail dot com> <alpine dot LSU dot 2 dot 11 dot 1604041342040 dot 13384 at t29 dot fhfr dot qr> <20160404120030 dot GD14122 at kam dot mff dot cuni dot cz>

On 4 April 2016 at 17:30, Jan Hubicka <hubicka@ucw.cz> wrote:
>> > Um not sure if I understood correctly.
>> > Do we want to constrain individual partition size by adding parameter
>> > lto-max-partition
>> > for balanced partitioning but not for -flto-partition=one
>> > case (since latter would also change semantics of =one) ?
>>
>> Yes, I think so.
>
> Yep, I agree.  Having partition one that produces multiple partitions doesn't seem to make much sense.
> Given that we have such not so predictable target specific limits on size of single translation unit
> we can handle, I suppose adding a resonable limit to the default balanced partitioning makes more sense.
> One can always push the behaviour you intend by setting max partitions to 1 (I suppose max size should
> have precedence over max partitions)
Thanks for the suggestions, I have updated the patch.
Is it OK if it passes bootstrap+test ?

Thanks,
Prathamesh
>
> Honza
>>
>> Richard.
>>
>> > Thanks,
>> > Prathamesh
>> > >
>> > > Richard.
>> > >
>> > > --
>> > > Richard Biener <rguenther@suse.de>
>> > > SUSE LINUX GmbH, GF: Felix Imendoerffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nuernberg)
>> >
>> >
>>
>> --
>> Richard Biener <rguenther@suse.de>
>> SUSE LINUX GmbH, GF: Felix Imendoerffer, Jane Smithard, Graham Norton, HRB 21284 (AG Nuernberg)

Attachment: patch-2.diff
Description: Text document

Attachment: ChangeLog
Description: Binary data

Follow-Ups:
- Re: [RFC] introduce --param max-lto-partition for having an upper bound on partition size
  - From: Jan Hubicka

References:
- [RFC] introduce --param max-lto-partition for having an upper bound on partition size
  - From: Prathamesh Kulkarni
- Re: [RFC] introduce --param max-lto-partition for having an upper bound on partition size
  - From: Richard Biener
- Re: [RFC] introduce --param max-lto-partition for having an upper bound on partition size
  - From: Prathamesh Kulkarni
- Re: [RFC] introduce --param max-lto-partition for having an upper bound on partition size
  - From: Richard Biener
- Re: [RFC] introduce --param max-lto-partition for having an upper bound on partition size
  - From: Prathamesh Kulkarni
- Re: [RFC] introduce --param max-lto-partition for having an upper bound on partition size
  - From: Richard Biener
- Re: [RFC] introduce --param max-lto-partition for having an upper bound on partition size
  - From: Jan Hubicka

Index Nav:	[Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav:	[Date Prev] [Date Next]	[Thread Prev] [Thread Next]