* [PATCH] Fix PR80334
@ 2017-04-06 12:40 Richard Biener
2017-04-06 20:45 ` Rainer Orth
0 siblings, 1 reply; 7+ messages in thread
From: Richard Biener @ 2017-04-06 12:40 UTC (permalink / raw)
To: gcc-patches
The following patch makes sure to preserve (mis-)alignment of memory
references when IVOPTs generates TARGET_MEM_REFs for them.
Bootstrapped on x86_64-unknown-linux-gnu, testing in progress.
Richard.
2017-04-06 Richard Biener <rguenther@suse.de>
PR tree-optimization/80334
* tree-ssa-loop-ivopts.c (rewrite_use_address): Properly
preserve alignment of accesses.
* g++.dg/torture/pr80334.C: New testcase.
Index: gcc/tree-ssa-loop-ivopts.c
===================================================================
--- gcc/tree-ssa-loop-ivopts.c (revision 246724)
+++ gcc/tree-ssa-loop-ivopts.c (working copy)
@@ -7396,7 +7396,11 @@ rewrite_use_address (struct ivopts_data
base_hint = var_at_stmt (data->current_loop, cand, use->stmt);
iv = var_at_stmt (data->current_loop, cand, use->stmt);
- ref = create_mem_ref (&bsi, TREE_TYPE (*use->op_p), &aff,
+ tree type = TREE_TYPE (*use->op_p);
+ unsigned int align = get_object_alignment (*use->op_p);
+ if (align != TYPE_ALIGN (type))
+ type = build_aligned_type (type, align);
+ ref = create_mem_ref (&bsi, type, &aff,
reference_alias_ptr_type (*use->op_p),
iv, base_hint, data->speed);
copy_ref_info (ref, *use->op_p);
Index: gcc/testsuite/g++.dg/torture/pr80334.C
===================================================================
--- gcc/testsuite/g++.dg/torture/pr80334.C (nonexistent)
+++ gcc/testsuite/g++.dg/torture/pr80334.C (working copy)
@@ -0,0 +1,18 @@
+// { dg-do run }
+
+struct A { alignas(16) char c; };
+struct B { A unpacked; char d; } __attribute__((packed));
+
+char x;
+
+int
+main()
+{
+ alignas(16) B b[3];
+ for (int i = 0; i < 3; i++) b[i].unpacked.c = 'a' + i;
+ for (int i = 0; i < 3; i++)
+ {
+ auto a = new A(b[i].unpacked);
+ x = a->c;
+ }
+}
^ permalink raw reply [flat|nested] 7+ messages in thread* Re: [PATCH] Fix PR80334
2017-04-06 12:40 [PATCH] Fix PR80334 Richard Biener
@ 2017-04-06 20:45 ` Rainer Orth
2017-04-07 6:49 ` Richard Biener
0 siblings, 1 reply; 7+ messages in thread
From: Rainer Orth @ 2017-04-06 20:45 UTC (permalink / raw)
To: Richard Biener; +Cc: gcc-patches
Hi Richard,
> The following patch makes sure to preserve (mis-)alignment of memory
> references when IVOPTs generates TARGET_MEM_REFs for them.
>
> Bootstrapped on x86_64-unknown-linux-gnu, testing in progress.
>
> Richard.
>
> 2017-04-06 Richard Biener <rguenther@suse.de>
>
> PR tree-optimization/80334
> * tree-ssa-loop-ivopts.c (rewrite_use_address): Properly
> preserve alignment of accesses.
>
> * g++.dg/torture/pr80334.C: New testcase.
the new testcase FAILs on 32-bit Solaris/SPARC:
+FAIL: g++.dg/torture/pr80334.C -O0 (test for excess errors)
+FAIL: g++.dg/torture/pr80334.C -O1 (test for excess errors)
+FAIL: g++.dg/torture/pr80334.C -O2 (test for excess errors)
+FAIL: g++.dg/torture/pr80334.C -O2 -flto (test for excess errors)
+FAIL: g++.dg/torture/pr80334.C -O2 -flto -flto-partition=none (test for exce
ss errors)
+FAIL: g++.dg/torture/pr80334.C -O3 -fomit-frame-pointer -funroll-loops -fpeel
-loops -ftracer -finline-functions (test for excess errors)
+FAIL: g++.dg/torture/pr80334.C -O3 -g (test for excess errors)
+FAIL: g++.dg/torture/pr80334.C -Os (test for excess errors)
Excess errors:
/vol/gcc/src/hg/trunk/local/gcc/testsuite/g++.dg/torture/pr80334.C:11:20: warning: requested alignment 16 is larger than 8 [-Wattributes]
Rainer
--
-----------------------------------------------------------------------------
Rainer Orth, Center for Biotechnology, Bielefeld University
^ permalink raw reply [flat|nested] 7+ messages in thread
* Re: [PATCH] Fix PR80334
2017-04-06 20:45 ` Rainer Orth
@ 2017-04-07 6:49 ` Richard Biener
2017-04-07 8:21 ` Rainer Orth
0 siblings, 1 reply; 7+ messages in thread
From: Richard Biener @ 2017-04-07 6:49 UTC (permalink / raw)
To: Rainer Orth; +Cc: gcc-patches
On Thu, 6 Apr 2017, Rainer Orth wrote:
> Hi Richard,
>
> > The following patch makes sure to preserve (mis-)alignment of memory
> > references when IVOPTs generates TARGET_MEM_REFs for them.
> >
> > Bootstrapped on x86_64-unknown-linux-gnu, testing in progress.
> >
> > Richard.
> >
> > 2017-04-06 Richard Biener <rguenther@suse.de>
> >
> > PR tree-optimization/80334
> > * tree-ssa-loop-ivopts.c (rewrite_use_address): Properly
> > preserve alignment of accesses.
> >
> > * g++.dg/torture/pr80334.C: New testcase.
>
> the new testcase FAILs on 32-bit Solaris/SPARC:
>
> +FAIL: g++.dg/torture/pr80334.C -O0 (test for excess errors)
> +FAIL: g++.dg/torture/pr80334.C -O1 (test for excess errors)
> +FAIL: g++.dg/torture/pr80334.C -O2 (test for excess errors)
> +FAIL: g++.dg/torture/pr80334.C -O2 -flto (test for excess errors)
> +FAIL: g++.dg/torture/pr80334.C -O2 -flto -flto-partition=none (test for exce
> ss errors)
> +FAIL: g++.dg/torture/pr80334.C -O3 -fomit-frame-pointer -funroll-loops -fpeel
> -loops -ftracer -finline-functions (test for excess errors)
> +FAIL: g++.dg/torture/pr80334.C -O3 -g (test for excess errors)
> +FAIL: g++.dg/torture/pr80334.C -Os (test for excess errors)
>
> Excess errors:
> /vol/gcc/src/hg/trunk/local/gcc/testsuite/g++.dg/torture/pr80334.C:11:20: warning: requested alignment 16 is larger than 8 [-Wattributes]
Any suggestion how to mitigate that? Possible solution includes
adding { target { ! ... } } to dg-do run.
Richard.
^ permalink raw reply [flat|nested] 7+ messages in thread* Re: [PATCH] Fix PR80334
2017-04-07 6:49 ` Richard Biener
@ 2017-04-07 8:21 ` Rainer Orth
2017-04-07 8:33 ` Richard Biener
0 siblings, 1 reply; 7+ messages in thread
From: Rainer Orth @ 2017-04-07 8:21 UTC (permalink / raw)
To: Richard Biener; +Cc: gcc-patches
Hi Richard,
> On Thu, 6 Apr 2017, Rainer Orth wrote:
>
>> Hi Richard,
>>
>> > The following patch makes sure to preserve (mis-)alignment of memory
>> > references when IVOPTs generates TARGET_MEM_REFs for them.
>> >
>> > Bootstrapped on x86_64-unknown-linux-gnu, testing in progress.
>> >
>> > Richard.
>> >
>> > 2017-04-06 Richard Biener <rguenther@suse.de>
>> >
>> > PR tree-optimization/80334
>> > * tree-ssa-loop-ivopts.c (rewrite_use_address): Properly
>> > preserve alignment of accesses.
>> >
>> > * g++.dg/torture/pr80334.C: New testcase.
>>
>> the new testcase FAILs on 32-bit Solaris/SPARC:
>>
>> +FAIL: g++.dg/torture/pr80334.C -O0 (test for excess errors)
>> +FAIL: g++.dg/torture/pr80334.C -O1 (test for excess errors)
>> +FAIL: g++.dg/torture/pr80334.C -O2 (test for excess errors)
>> +FAIL: g++.dg/torture/pr80334.C -O2 -flto (test for excess errors)
>> +FAIL: g++.dg/torture/pr80334.C -O2 -flto -flto-partition=none (test for
>> exce
>> ss errors)
>> +FAIL: g++.dg/torture/pr80334.C -O3 -fomit-frame-pointer -funroll-loops
>> -fpeel
>> -loops -ftracer -finline-functions (test for excess errors)
>> +FAIL: g++.dg/torture/pr80334.C -O3 -g (test for excess errors)
>> +FAIL: g++.dg/torture/pr80334.C -Os (test for excess errors)
>>
>> Excess errors:
>> /vol/gcc/src/hg/trunk/local/gcc/testsuite/g++.dg/torture/pr80334.C:11:20:
>> warning: requested alignment 16 is larger than 8 [-Wattributes]
>
> Any suggestion how to mitigate that? Possible solution includes
> adding { target { ! ... } } to dg-do run.
No idea. However, according to gcc-testresults there are other
failures: s390-ibm-inux-gnu and s390x-ibm-linux-gnu so far.
This might argue against just excluding a random list of failing targets.
Rainer
--
-----------------------------------------------------------------------------
Rainer Orth, Center for Biotechnology, Bielefeld University
^ permalink raw reply [flat|nested] 7+ messages in thread* Re: [PATCH] Fix PR80334
2017-04-07 8:21 ` Rainer Orth
@ 2017-04-07 8:33 ` Richard Biener
2017-04-07 8:41 ` Rainer Orth
0 siblings, 1 reply; 7+ messages in thread
From: Richard Biener @ 2017-04-07 8:33 UTC (permalink / raw)
To: Rainer Orth; +Cc: gcc-patches
On Fri, 7 Apr 2017, Rainer Orth wrote:
> Hi Richard,
>
> > On Thu, 6 Apr 2017, Rainer Orth wrote:
> >
> >> Hi Richard,
> >>
> >> > The following patch makes sure to preserve (mis-)alignment of memory
> >> > references when IVOPTs generates TARGET_MEM_REFs for them.
> >> >
> >> > Bootstrapped on x86_64-unknown-linux-gnu, testing in progress.
> >> >
> >> > Richard.
> >> >
> >> > 2017-04-06 Richard Biener <rguenther@suse.de>
> >> >
> >> > PR tree-optimization/80334
> >> > * tree-ssa-loop-ivopts.c (rewrite_use_address): Properly
> >> > preserve alignment of accesses.
> >> >
> >> > * g++.dg/torture/pr80334.C: New testcase.
> >>
> >> the new testcase FAILs on 32-bit Solaris/SPARC:
> >>
> >> +FAIL: g++.dg/torture/pr80334.C -O0 (test for excess errors)
> >> +FAIL: g++.dg/torture/pr80334.C -O1 (test for excess errors)
> >> +FAIL: g++.dg/torture/pr80334.C -O2 (test for excess errors)
> >> +FAIL: g++.dg/torture/pr80334.C -O2 -flto (test for excess errors)
> >> +FAIL: g++.dg/torture/pr80334.C -O2 -flto -flto-partition=none (test for
> >> exce
> >> ss errors)
> >> +FAIL: g++.dg/torture/pr80334.C -O3 -fomit-frame-pointer -funroll-loops
> >> -fpeel
> >> -loops -ftracer -finline-functions (test for excess errors)
> >> +FAIL: g++.dg/torture/pr80334.C -O3 -g (test for excess errors)
> >> +FAIL: g++.dg/torture/pr80334.C -Os (test for excess errors)
> >>
> >> Excess errors:
> >> /vol/gcc/src/hg/trunk/local/gcc/testsuite/g++.dg/torture/pr80334.C:11:20:
> >> warning: requested alignment 16 is larger than 8 [-Wattributes]
> >
> > Any suggestion how to mitigate that? Possible solution includes
> > adding { target { ! ... } } to dg-do run.
>
> No idea. However, according to gcc-testresults there are other
> failures: s390-ibm-inux-gnu and s390x-ibm-linux-gnu so far.
>
> This might argue against just excluding a random list of failing targets.
Hmm. Does using __BIGGEST_ALIGNMENT__, thus
int
main()
{
alignas(__BIGGEST_ALIGNMENT__) B b[3];
...
work for you?
Richard.
^ permalink raw reply [flat|nested] 7+ messages in thread* Re: [PATCH] Fix PR80334
2017-04-07 8:33 ` Richard Biener
@ 2017-04-07 8:41 ` Rainer Orth
2017-04-07 8:45 ` Richard Biener
0 siblings, 1 reply; 7+ messages in thread
From: Rainer Orth @ 2017-04-07 8:41 UTC (permalink / raw)
To: Richard Biener; +Cc: gcc-patches
Hi Richard,
>> > Any suggestion how to mitigate that? Possible solution includes
>> > adding { target { ! ... } } to dg-do run.
>>
>> No idea. However, according to gcc-testresults there are other
>> failures: s390-ibm-inux-gnu and s390x-ibm-linux-gnu so far.
>>
>> This might argue against just excluding a random list of failing targets.
>
> Hmm. Does using __BIGGEST_ALIGNMENT__, thus
>
> int
> main()
> {
> alignas(__BIGGEST_ALIGNMENT__) B b[3];
> ...
>
> work for you?
it does: the test now PASSes on sparc-sun-solaris2.12 and continues to
do so in i386-pc-solaris2.12, both 32 and 64-bit.
Thanks.
Rainer
--
-----------------------------------------------------------------------------
Rainer Orth, Center for Biotechnology, Bielefeld University
^ permalink raw reply [flat|nested] 7+ messages in thread* Re: [PATCH] Fix PR80334
2017-04-07 8:41 ` Rainer Orth
@ 2017-04-07 8:45 ` Richard Biener
0 siblings, 0 replies; 7+ messages in thread
From: Richard Biener @ 2017-04-07 8:45 UTC (permalink / raw)
To: Rainer Orth; +Cc: gcc-patches
On Fri, 7 Apr 2017, Rainer Orth wrote:
> Hi Richard,
>
> >> > Any suggestion how to mitigate that? Possible solution includes
> >> > adding { target { ! ... } } to dg-do run.
> >>
> >> No idea. However, according to gcc-testresults there are other
> >> failures: s390-ibm-inux-gnu and s390x-ibm-linux-gnu so far.
> >>
> >> This might argue against just excluding a random list of failing targets.
> >
> > Hmm. Does using __BIGGEST_ALIGNMENT__, thus
> >
> > int
> > main()
> > {
> > alignas(__BIGGEST_ALIGNMENT__) B b[3];
> > ...
> >
> > work for you?
>
> it does: the test now PASSes on sparc-sun-solaris2.12 and continues to
> do so in i386-pc-solaris2.12, both 32 and 64-bit.
Tested on x86_64-unknwon-linux-gnu, applied.
Richard.
2017-04-07 Richard Biener <rguenther@suse.de>
PR tree-optimization/80334
* g++.dg/torture/pr80334.C: Use __BIGGEST_ALIGNMENT__ for
alignas on stack.
Index: gcc/testsuite/g++.dg/torture/pr80334.C
===================================================================
--- gcc/testsuite/g++.dg/torture/pr80334.C (revision 246752)
+++ gcc/testsuite/g++.dg/torture/pr80334.C (working copy)
@@ -8,7 +8,7 @@ char x;
int
main()
{
- alignas(16) B b[3];
+ alignas(__BIGGEST_ALIGNMENT__) B b[3];
for (int i = 0; i < 3; i++) b[i].unpacked.c = 'a' + i;
for (int i = 0; i < 3; i++)
{
^ permalink raw reply [flat|nested] 7+ messages in thread
end of thread, other threads:[~2017-04-07 8:45 UTC | newest]
Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2017-04-06 12:40 [PATCH] Fix PR80334 Richard Biener
2017-04-06 20:45 ` Rainer Orth
2017-04-07 6:49 ` Richard Biener
2017-04-07 8:21 ` Rainer Orth
2017-04-07 8:33 ` Richard Biener
2017-04-07 8:41 ` Rainer Orth
2017-04-07 8:45 ` Richard Biener
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).