Concepts-checking and other kinds of early tsubsting may often take
place while location wrappers are suppressed, e.g. because we've
triggered template instantiation within template parameter lists.
With that, exprs that are usually wrapped by VIEW_CONVERT_EXPRs
location wrappers may end up wrappe
On Sun, Dec 30, 2018 at 1:06 AM Max Filippov wrote:
>
> Xtensa architecture is not affected by speculation.
>
> gcc/
> 2018-12-30 Max Filippov
>
> * config/xtensa/xtensa.c (TARGET_HAVE_SPECULATION_SAFE_VALUE):
> Define to speculation_safe_value_not_needed.
Approved.
Just a trivial cleanup from Martin's work.
if (x == CONST)
do something;
else
return value;
Turns into
if (x != CONST)
return value;
do something;
Again no behavior changes, but makes it easier to find the nuggets in
the rest of Martin's work.
Bootstrapped, regression tested on x86_64
This is another chunk of Martin's work that I've extracted. It removes
the "type" and "fuzzy" parameters we pass to get_range_strlen and
replaces them with a strlen_range_kind enum. Mostly this cleans up the
API and avoids the possibility of bogus combinations of "type" and "fuzzy".
It has one e
On Sun, Dec 30, 2018 at 09:37:20AM -0800, Steve Kargl wrote:
> On Sun, Dec 30, 2018 at 06:10:14PM +0100, Thomas Koenig wrote:
>
> > + {
> > + tree present;
> > + tree type;
> > +
> > + type = TREE_TYPE (maskse.expr);
> > + present = gfc_conv_expr_present (maskexpr->symtree->n.sym
On Sun, Dec 30, 2018 at 09:32:43AM -0800, H.J. Lu wrote:
> On Sun, Dec 30, 2018 at 9:09 AM Steve Kargl
> wrote:
> >
> > On Sun, Dec 30, 2018 at 08:06:40AM -0800, H.J. Lu wrote:
> > > >
> > > > OK Steve, thanks.
> > >
> > > The test fails on Linux/x86:
> > >
> > > https://gcc.gnu.org/bugzilla/show_
On Sun, Dec 30, 2018 at 06:10:14PM +0100, Thomas Koenig wrote:
> Index: gcc/fortran/trans-expr.c
> ===
> --- gcc/fortran/trans-expr.c (Revision 267347)
> +++ gcc/fortran/trans-expr.c (Arbeitskopie)
> @@ -5760,17 +5760,21 @@ gfc_conv
On Sun, Dec 30, 2018 at 9:09 AM Steve Kargl
wrote:
>
> On Sun, Dec 30, 2018 at 08:06:40AM -0800, H.J. Lu wrote:
> > >
> > > OK Steve, thanks.
> >
> > The test fails on Linux/x86:
> >
> > https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88639
> >
>
> The test works on both i586-*-freebsd and x86_64-*-f
On Sun, Dec 30, 2018 at 08:06:40AM -0800, H.J. Lu wrote:
> >
> > OK Steve, thanks.
>
> The test fails on Linux/x86:
>
> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=88639
>
The test works on both i586-*-freebsd and x86_64-*-freebsd.
What does
--
Steve
k;
}
[hjl@gnu-cfl-1 pr87007]$ make test2.s
/export/build/gnu/tools-build/gcc-debug/build-x86_64-linux/gcc/xgcc
-B/export/build/gnu/tools-build/gcc-debug/build-x86_64-linux/gcc/ -O2
-mavx -S test2.i
[hjl@gnu-cfl-1 pr87007]$ cat test2.s
.file "test2.i"
.text
.p2align 4
.globl foo
.type foo, @function
foo:
.LFB0:
.cfi_startproc
vmovss .LC0(%rip), %xmm0
cmpl $1, %edi
je .L15
vmovss .LC1(%rip), %xmm0
cmpl $4, %edi
je .L15
vxorps %xmm0, %xmm0, %xmm0
testl %edi, %edi
je .L3
leal -1(%rdi), %ecx
xorl %eax, %eax
jmp .L6
.p2align 4,,10
.p2align 3
.L17:
vcvtsi2ss i1(,%rax,4), %xmm0, %xmm1
leaq 1(%rax), %rdx
vmovss %xmm1, f1(,%rax,4)
cmpq %rcx, %rax
je .L3
.L9:
movq %rdx, %rax
.L6:
cmpl $100, (%rsi,%rax,4)
jg .L17
vcvtsi2ss i2(,%rax,4), %xmm0, %xmm1
leaq 1(%rax), %rdx
vmovss %xmm1, f2(,%rax,4)
cmpq %rcx, %rax
jne .L9
.L3:
vcvtsi2ss %edi, %xmm0, %xmm0
.L15:
ret
.cfi_endproc
.LFE0:
.size foo, .-foo
.section .rodata.cst4,"aM",@progbits,4
.align 4
.LC0:
.long 1065353216
.align 4
.LC1:
.long 1084227584
.ident "GCC: (GNU) 9.0.0 20181230 (experimental)"
.section .note.GNU-stack,"",@progbits
[hjl@gnu-cfl-1 pr87007]$
The placement is optimal.
--
H.J.
On Wed, Nov 28, 2018 at 12:17 PM Jeff Law wrote:
>
> On 11/28/18 12:48 PM, H.J. Lu wrote:
> > On Mon, Nov 5, 2018 at 7:29 AM Jan Hubicka wrote:
> >>
> >>> On 11/5/18 7:21 AM, Jan Hubicka wrote:
> >
> > Did you mean "the nearest common dominator"?
>
> If the nearest common domina
This PR points out that while we are able to deduce the template arguments
for A{}, we fail for A().
For A{}, the deduction happens in finish_compound_literal:
2789 if (tree anode = type_uses_auto (type))
2790 if (CLASS_PLACEHOLDER_TEMPLATE (anode))
2791 {
2792 type = do_auto_d
On Fri, Dec 28, 2018 at 4:38 PM Jerry DeLisle wrote:
>
> On 12/28/18 10:43 AM, Steve Kargl wrote:
> > Ping.
> >
> > On Mon, Dec 24, 2018 at 11:59:50AM -0800, Steve Kargl wrote:
> >> All,
> >>
> >> The IEEE modules and -ffpe-trap are to some extent orthogonal
> >> features of gfortran. Unfortunate
There have been some ICEs in the past for msp430-elf with the large
memory model (20-bit pointers), caused by SET_{DECL,TYPE}_ALIGN being called
with an argument which resolves to 20 i.e. POINTER_SIZE,
or the default value of TARGET_VTABLE_ENTRY_ALIGN (which is POINTER_SIZE).
The attached patch ad
Xtensa architecture is not affected by speculation.
gcc/
2018-12-30 Max Filippov
* config/xtensa/xtensa.c (TARGET_HAVE_SPECULATION_SAFE_VALUE):
Define to speculation_safe_value_not_needed.
---
gcc/config/xtensa/xtensa.c | 3 +++
1 file changed, 3 insertions(+)
diff --git a/gc
15 matches
Mail list logo
