This is the mail archive of the
gcc-patches@gcc.gnu.org
mailing list for the GCC project.
Re: [PATCH] varasm: Propagate litpool decl alignment to generated RTX.
- From: Andreas Krebbel <krebbel at linux dot vnet dot ibm dot com>
- To: Richard Biener <richard dot guenther at gmail dot com>
- Cc: GCC Patches <gcc-patches at gcc dot gnu dot org>
- Date: Thu, 22 Dec 2016 13:58:10 +0100
- Subject: Re: [PATCH] varasm: Propagate litpool decl alignment to generated RTX.
- Authentication-results: sourceware.org; auth=none
- References: <20161216202922.14425-1-krebbel@linux.vnet.ibm.com> <CAFiYyc16HhrW-p6T9C840PeKWVHJy=QxoBMHUMpizvMfiG6dNw@mail.gmail.com>
On 12/20/2016 11:38 AM, Richard Biener wrote:
> On Fri, Dec 16, 2016 at 9:29 PM, Andreas Krebbel
> <krebbel@linux.vnet.ibm.com> wrote:
>> When pushing a value into the literal pool the resulting decl might
>> get a higher alignment than the original expression depending on how a
>> target defines CONSTANT_ALIGNMENT. Generating an RTX for the constant
>> pool access we currently use the alignment from the original
>> expression. Changed with the attached patch.
>
> And it might be even smaller alignment... or do we not allow that?
I did assume that this is not supposed to happen. Adding an assertion triggering in that case
survived bootstrap and testsuite. s390x only. It basically boils down to whether align_variable and
set_mem_attributes/get_object_alignment come to different conclusions about the alignment starting
at either the var decl or the original expression.
...
>> Bootstrapped and regtested on x86_64 and s390x.
>>
>> No regressions.
>>
>> Ok?
>
> Ok.
>
> Richard.
Ok for GCC 6 branch as well?
-Andreas-