Re: Base subreg rules on REGMODE_NATURAL_SIZE rather than UNITS_PER_WORD
On 11/06/2017 07:53 AM, Richard Sandiford wrote: >>> 2017-09-18 Richard Sandiford >>> Alan Hayward >>> David Sherwood >>> >>> gcc/ >>> * doc/rtl.texi: Rewrite the subreg rules so that they partition >>> the inner register into REGMODE_NATURAL_SIZE bytes rather than >>> UNITS_PER_WORD bytes. >>> * emit-rtl.c (validate_subreg): Divide subregs into blocks >>> based on REGMODE_NATURAL_SIZE of the inner mode. >>> (gen_lowpart_common): Split the SCALAR_FLOAT_MODE_P and >>> !SCALAR_FLOAT_MODE_P cases. Use REGMODE_NATURAL_SIZE for the latter. >>> * expr.c (store_constructor): Use REGMODE_NATURAL_SIZE to test >>> whether something is likely to occupy more than one register. >> OK. Though I must admit, I expected a larger change after reading the >> intro. > > Heh. As it turned out I did find one more place that needed to change > (lowpart_bit_field_p). Maybe there'll be more -- e.g. it'd probably > make sense to make lower-subreg.c split based on this instead of > UNITS_PER_WORD. > > How about this? Tested as before. > > Thanks, > Richard > > > 2017-11-06 Richard Sandiford > Alan Hayward > David Sherwood > > gcc/ > * doc/rtl.texi: Rewrite the subreg rules so that they partition > the inner register into REGMODE_NATURAL_SIZE bytes rather than > UNITS_PER_WORD bytes. > * emit-rtl.c (validate_subreg): Divide subregs into blocks > based on REGMODE_NATURAL_SIZE of the inner mode. > (gen_lowpart_common): Split the SCALAR_FLOAT_MODE_P and > !SCALAR_FLOAT_MODE_P cases. Use REGMODE_NATURAL_SIZE for the latter. > * expmed.c (lowpart_bit_field_p): Divide the value up into > chunks of REGMODE_NATURAL_SIZE rather than UNITS_PER_WORD. > * expr.c (store_constructor): Use REGMODE_NATURAL_SIZE to test > whether something is likely to occupy more than one register. OK. I still expect more fallout though :-) Jeff
Re: Base subreg rules on REGMODE_NATURAL_SIZE rather than UNITS_PER_WORD
Jeff Law writes:
> On 09/18/2017 05:26 AM, Richard Sandiford wrote:
>> Originally subregs operated at the word level and subreg offsets
>> were measured in words. The offset units were later changed from
>> words to bytes (SUBREG_WORD became SUBREG_BYTE), but the fundamental
>> assumption that subregs should operate at the word level remained.
>> Whether (subreg:M1 (reg:M2 R2) N) is well-formed depended on the
>> way that M1 and M2 partitioned into words and whether the subword
>> part of N represented a lowpart. However, some questions depended
>> instead on the macro REGMODE_NATURAL_SIZE, which was introduced
>> as part of the patch that moved from SUBREG_WORD to SUBREG_BYTE.
>> It is used to decide whether setting (subreg:M1 (reg:M2 R2) N)
>> clobbers all of R2 or just part of it (df_read_modify_subreg).
>>
>> Using words doesn't really make sense for modern vector
>> architectures. Vector registers are usually bigger than
>> a word and:
>>
>> (a) setting the scalar lowpart of them usually clobbers the
>> rest of the register (contrary to the subreg rules,
>> where only the containing words should be clobbered).
>>
>> (b) high words of vector registers are often not independently
>> addressable, even though that's what the subreg rules expect.
>>
>> This patch therefore uses REGMODE_NATURAL_SIZE instead of
>> UNITS_PER_WORD to determine the size of the independently
>> addressable blocks in an inner register.
>>
>> This is needed for SVE because the number of words in a vector
>> mode isn't known at compile time, so isn't a sensible basis
>> for calculating the number of registers.
>>
>> The only existing port to define REGMODE_NATURAL_SIZE is
>> 64-bit SPARC, where FP registers are 32 bits. (This is the
>> opposite of the use case for SVE, since the natural division
>> is smaller than a word.) I compiled the testsuite before and
>> after the patch for sparc64-linux-gnu and the only test whose
>> assembly changed was g++.dg/debug/pr65678.C, where the order
>> of two independent stores was reversed and where a different
>> register was picked for one pseudo. The new code was
>> otherwise equivalent to the old code.
>>
>> Tested on aarch64-linux-gnu, x86_64-linux-gnu and powerpc64le-linux-gnu.
>> Also tested by comparing the testsuite assembly output on at least one
>> target per CPU directory, with only the SPARC differences just mentioned.
>> OK to install?
>>
>> Richard
>>
>>
>> 2017-09-18 Richard Sandiford
>> Alan Hayward
>> David Sherwood
>>
>> gcc/
>> * doc/rtl.texi: Rewrite the subreg rules so that they partition
>> the inner register into REGMODE_NATURAL_SIZE bytes rather than
>> UNITS_PER_WORD bytes.
>> * emit-rtl.c (validate_subreg): Divide subregs into blocks
>> based on REGMODE_NATURAL_SIZE of the inner mode.
>> (gen_lowpart_common): Split the SCALAR_FLOAT_MODE_P and
>> !SCALAR_FLOAT_MODE_P cases. Use REGMODE_NATURAL_SIZE for the latter.
>> * expr.c (store_constructor): Use REGMODE_NATURAL_SIZE to test
>> whether something is likely to occupy more than one register.
> OK. Though I must admit, I expected a larger change after reading the
> intro.
Heh. As it turned out I did find one more place that needed to change
(lowpart_bit_field_p). Maybe there'll be more -- e.g. it'd probably
make sense to make lower-subreg.c split based on this instead of
UNITS_PER_WORD.
How about this? Tested as before.
Thanks,
Richard
2017-11-06 Richard Sandiford
Alan Hayward
David Sherwood
gcc/
* doc/rtl.texi: Rewrite the subreg rules so that they partition
the inner register into REGMODE_NATURAL_SIZE bytes rather than
UNITS_PER_WORD bytes.
* emit-rtl.c (validate_subreg): Divide subregs into blocks
based on REGMODE_NATURAL_SIZE of the inner mode.
(gen_lowpart_common): Split the SCALAR_FLOAT_MODE_P and
!SCALAR_FLOAT_MODE_P cases. Use REGMODE_NATURAL_SIZE for the latter.
* expmed.c (lowpart_bit_field_p): Divide the value up into
chunks of REGMODE_NATURAL_SIZE rather than UNITS_PER_WORD.
* expr.c (store_constructor): Use REGMODE_NATURAL_SIZE to test
whether something is likely to occupy more than one register.
Index: gcc/doc/rtl.texi
===
--- gcc/doc/rtl.texi2017-11-06 14:44:22.098357016 +
+++ gcc/doc/rtl.texi2017-11-06 14:49:48.804883947 +
@@ -1925,19 +1925,32 @@ false.
When @var{m1} is at least as narrow as @var{m2} the @code{subreg}
expression is called @dfn{normal}.
+@findex REGMODE_NATURAL_SIZE
Normal @code{subreg}s restrict consideration to certain bits of
-@var{reg}. There are two cases. If @var{m1} is smaller than a word,
-the @code{subreg} refers to the least-significant part (or
-@dfn{lowpart}) of one word of @var{reg}. If @var{m1} is word-sized or
-greater, the @code{subreg} refers to one or more
Re: Base subreg rules on REGMODE_NATURAL_SIZE rather than UNITS_PER_WORD
On 09/18/2017 05:26 AM, Richard Sandiford wrote: > Originally subregs operated at the word level and subreg offsets > were measured in words. The offset units were later changed from > words to bytes (SUBREG_WORD became SUBREG_BYTE), but the fundamental > assumption that subregs should operate at the word level remained. > Whether (subreg:M1 (reg:M2 R2) N) is well-formed depended on the > way that M1 and M2 partitioned into words and whether the subword > part of N represented a lowpart. However, some questions depended > instead on the macro REGMODE_NATURAL_SIZE, which was introduced > as part of the patch that moved from SUBREG_WORD to SUBREG_BYTE. > It is used to decide whether setting (subreg:M1 (reg:M2 R2) N) > clobbers all of R2 or just part of it (df_read_modify_subreg). > > Using words doesn't really make sense for modern vector > architectures. Vector registers are usually bigger than > a word and: > > (a) setting the scalar lowpart of them usually clobbers the > rest of the register (contrary to the subreg rules, > where only the containing words should be clobbered). > > (b) high words of vector registers are often not independently > addressable, even though that's what the subreg rules expect. > > This patch therefore uses REGMODE_NATURAL_SIZE instead of > UNITS_PER_WORD to determine the size of the independently > addressable blocks in an inner register. > > This is needed for SVE because the number of words in a vector > mode isn't known at compile time, so isn't a sensible basis > for calculating the number of registers. > > The only existing port to define REGMODE_NATURAL_SIZE is > 64-bit SPARC, where FP registers are 32 bits. (This is the > opposite of the use case for SVE, since the natural division > is smaller than a word.) I compiled the testsuite before and > after the patch for sparc64-linux-gnu and the only test whose > assembly changed was g++.dg/debug/pr65678.C, where the order > of two independent stores was reversed and where a different > register was picked for one pseudo. The new code was > otherwise equivalent to the old code. > > Tested on aarch64-linux-gnu, x86_64-linux-gnu and powerpc64le-linux-gnu. > Also tested by comparing the testsuite assembly output on at least one > target per CPU directory, with only the SPARC differences just mentioned. > OK to install? > > Richard > > > 2017-09-18 Richard Sandiford > Alan Hayward > David Sherwood > > gcc/ > * doc/rtl.texi: Rewrite the subreg rules so that they partition > the inner register into REGMODE_NATURAL_SIZE bytes rather than > UNITS_PER_WORD bytes. > * emit-rtl.c (validate_subreg): Divide subregs into blocks > based on REGMODE_NATURAL_SIZE of the inner mode. > (gen_lowpart_common): Split the SCALAR_FLOAT_MODE_P and > !SCALAR_FLOAT_MODE_P cases. Use REGMODE_NATURAL_SIZE for the latter. > * expr.c (store_constructor): Use REGMODE_NATURAL_SIZE to test > whether something is likely to occupy more than one register. OK. Though I must admit, I expected a larger change after reading the intro. jeff
