Re: [PATCH] Fix aarch64 bootstrap (pr69416)

2016-01-25 Thread Christophe Lyon
On 22 January 2016 at 18:07, Richard Henderson wrote: > The bare CONST_INT inside the CCmode IF_THEN_ELSE is causing combine to make > incorrect simplifications. At this stage it feels safer to wrap the > CONST_INT inside of an UNSPEC than make more generic changes to combine. >

Re: [PATCH] Fix aarch64 bootstrap (pr69416)

2016-01-25 Thread Richard Henderson
On 01/25/2016 05:28 AM, Christophe Lyon wrote: > After this, I'm seeing this test now FAILs: > gcc.target/aarch64/ccmp_1.c scan-assembler adds\t That test case is badly written. In addition to that one, several of the other failures that I see within that file are simply equally optimal

Re: [PATCH] Fix aarch64 bootstrap (pr69416)

2016-01-25 Thread Wilco Dijkstra
Richard Henderson wrote: > On 01/25/2016 05:28 AM, Christophe Lyon wrote: > > After this, I'm seeing this test now FAILs: > > gcc.target/aarch64/ccmp_1.c scan-assembler adds\t > > That test case is badly written. In addition to that one, several of the > other > failures that I see within that

[PATCH] Fix aarch64 bootstrap (pr69416)

2016-01-22 Thread Richard Henderson
The bare CONST_INT inside the CCmode IF_THEN_ELSE is causing combine to make incorrect simplifications. At this stage it feels safer to wrap the CONST_INT inside of an UNSPEC than make more generic changes to combine. But we should definitely investigate combine's CCmode issues for gcc7.

Re: [PATCH] Fix aarch64 bootstrap (pr69416)

2016-01-22 Thread Richard Earnshaw (lists)
On 22/01/16 17:07, Richard Henderson wrote: > The bare CONST_INT inside the CCmode IF_THEN_ELSE is causing combine to > make incorrect simplifications. At this stage it feels safer to wrap > the CONST_INT inside of an UNSPEC than make more generic changes to > combine. > > But we should