Hi Eric,
On Mon, Nov 07, 2016 at 10:44:44AM +0100, Eric Botcazou wrote:
> Tested on PowerPC64/Linux, OK for the mainline?
>
>
>* config/rs6000/rs6000.c (rs6000_emit_move): Also use a TOC reference
> after forcing to constant memory when the code model is medium.
Sorry I lost tra
> Now you don't need to have a special pool to call create_TOC_reference, you
> can call it for regular TOC references as well, as done a few lines above:
>
> /* If this is a SYMBOL_REF that refers to a constant pool entry,
>and we have put it in the TOC, we just need to make a TOC-r
> I don't fully understand what you mean. This code was created for
> PR65810, if that helps?
OK, let's turn it into "mode" then, this doesn't change anything.
--
Eric Botcazou
On Tue, Oct 18, 2016 at 08:37:47PM +0200, Eric Botcazou wrote:
> > We need to pass the mode of the actual datum we would put in the TOC to
> > the use_toc_relative_ref function, not the mode of its address.
>
> Right, but this mode is not "mode", the TOC contains only Pmode entries if
> the
> sp
> We need to pass the mode of the actual datum we would put in the TOC to
> the use_toc_relative_ref function, not the mode of its address.
Right, but this mode is not "mode", the TOC contains only Pmode entries if the
special constant pool is excluded.
--
Eric Botcazou
On Tue, Oct 18, 2016 at 01:09:24PM +0200, Eric Botcazou wrote:
> > > No, "mode" is the mode of the MEM, not that of the SYMBOL_REF.
> >
> > I still don't see it, could you explain a bit more?
>
> MODE is the mode of operands[1] before:
>
> operands[1] = force_const_mem (mode, operands[1]
> > No, "mode" is the mode of the MEM, not that of the SYMBOL_REF.
>
> I still don't see it, could you explain a bit more?
MODE is the mode of operands[1] before:
operands[1] = force_const_mem (mode, operands[1]);
and after. But the test is on the address of the MEM, not on the MEM i
[ sorry for losing track of this patch ]
On Sun, Oct 09, 2016 at 10:32:51AM +0200, Eric Botcazou wrote:
> > Use "mode" instead of "Pmode" here?
>
> No, "mode" is the mode of the MEM, not that of the SYMBOL_REF.
I still don't see it, could you explain a bit more?
Segher
> Why does this not fail on darwin? It can reach this code afaics, and
> it has ASM_OUTPUT_SPECIAL_POOL_ENTRY_P always 0.
Probably because Darwin doesn't use the TOC at all.
> Use "mode" instead of "Pmode" here?
No, "mode" is the mode of the MEM, not that of the SYMBOL_REF.
--
Eric Botcazou
Hi!
On Sat, Oct 08, 2016 at 07:53:42PM +0200, Eric Botcazou wrote:
> this fixes reload failures in 64-bit mode with -mcmodel=medium for targets
> that define ASM_OUTPUT_SPECIAL_POOL_ENTRY_P to 0; that's the case for the
> VxWorks port we'll submit because of the compatibility with the system
>
Hi,
this fixes reload failures in 64-bit mode with -mcmodel=medium for targets
that define ASM_OUTPUT_SPECIAL_POOL_ENTRY_P to 0; that's the case for the
VxWorks port we'll submit because of the compatibility with the system linker,
which rejects .got sections in object files.
When reload ends
11 matches
Mail list logo
