rs6000: add option -mblock-ops-unaligned-vsx

This option is mostly being added to provide -mno-block-ops-unaligned-vsx.
The default is set the same as -mefficient-unaligned-vsx. This option will
control the use of unaligned VSX loads/stores in the inline expansion
of memcpy() and memmove(). The use case for this would be if you're
compiling code that is doing a memcpy to memory mapped device memory
that is cache-inhibited. On some powerpc processors this requires the
unaligned vsx ops to be emulated by the kernel which is very slow.

gcc/ChangeLog:

	* config/rs6000/rs6000.c (rs6000_option_override_internal):
	Set the default value for -mblock-ops-unaligned-vsx.
	* config/rs6000/rs6000.opt: Add -mblock-ops-unaligned-vsx.
	* doc/invoke.texi: Document -mblock-ops-unaligned-vsx.
This commit is contained in:
Aaron Sawdey 2020-07-24 14:13:48 -05:00
parent 2e764ae1d6
commit e398fcb9ce
3 changed files with 22 additions and 0 deletions

View File

@ -3979,6 +3979,14 @@ rs6000_option_override_internal (bool global_init_p)
}
}
if (!(rs6000_isa_flags_explicit & OPTION_MASK_BLOCK_OPS_UNALIGNED_VSX))
{
if (TARGET_EFFICIENT_UNALIGNED_VSX)
rs6000_isa_flags |= OPTION_MASK_BLOCK_OPS_UNALIGNED_VSX;
else
rs6000_isa_flags &= ~OPTION_MASK_BLOCK_OPS_UNALIGNED_VSX;
}
/* Use long double size to select the appropriate long double. We use
TYPE_PRECISION to differentiate the 3 different long double types. We map
128 into the precision used for TFmode. */
@ -23167,6 +23175,8 @@ struct rs6000_opt_mask {
static struct rs6000_opt_mask const rs6000_opt_masks[] =
{
{ "altivec", OPTION_MASK_ALTIVEC, false, true },
{ "block-ops-unaligned-vsx", OPTION_MASK_BLOCK_OPS_UNALIGNED_VSX,
false, true },
{ "cmpb", OPTION_MASK_CMPB, false, true },
{ "crypto", OPTION_MASK_CRYPTO, false, true },
{ "direct-move", OPTION_MASK_DIRECT_MOVE, false, true },

View File

@ -324,6 +324,10 @@ mblock-move-inline-limit=
Target Report Var(rs6000_block_move_inline_limit) Init(0) RejectNegative Joined UInteger Save
Max number of bytes to move inline.
mblock-ops-unaligned-vsx
Target Report Mask(BLOCK_OPS_UNALIGNED_VSX) Var(rs6000_isa_flags)
Generate unaligned VSX load/store for inline expansion of memcpy/memmove.
mblock-compare-inline-limit=
Target Report Var(rs6000_block_compare_inline_limit) Init(63) RejectNegative Joined UInteger Save
Max number of bytes to compare without loops.

View File

@ -1182,6 +1182,7 @@ See RS/6000 and PowerPC Options.
-mblock-move-inline-limit=@var{num} @gol
-mblock-compare-inline-limit=@var{num} @gol
-mblock-compare-inline-loop-limit=@var{num} @gol
-mno-block-ops-unaligned-vsx @gol
-mstring-compare-inline-limit=@var{num} @gol
-misel -mno-isel @gol
-mvrsave -mno-vrsave @gol
@ -27023,6 +27024,13 @@ store instructions when the option @option{-mcpu=future} is used.
@opindex mno-mma
Generate (do not generate) the MMA instructions when the option
@option{-mcpu=future} is used.
@item -mblock-ops-unaligned-vsx
@itemx -mno-block-ops-unaligned-vsx
@opindex block-ops-unaligned-vsx
@opindex no-block-ops-unaligned-vsx
Generate (do not generate) unaligned vsx loads and stores for
inline expansion of @code{memcpy} and @code{memmove}.
@end table
@node RX Options