* [PATCH v1] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32
@ 2023-06-14 0:58 pan2.li
2023-06-14 1:07 ` juzhe.zhong
` (2 more replies)
0 siblings, 3 replies; 11+ messages in thread
From: pan2.li @ 2023-06-14 0:58 UTC (permalink / raw)
To: gcc-patches
Cc: juzhe.zhong, rdapp.gcc, jeffreyalaw, pan2.li, yanzhang.wang, kito.cheng
From: Pan Li <pan2.li@intel.com>
This patch would like to fix one bug exported by RV32 test case
multiple_rgroup_run-2.c. The mask should be restricted by elen in
vector, and the condition between the vmv.s.x and the vmv.v.x should
take inner_bits_size rather than constants.
Passed both the rv32 and rv64 riscv/rvv tests.
Signed-off-by: Pan Li <pan2.li@intel.com>
gcc/ChangeLog:
* config/riscv/riscv-v.cc (rvv_builder::get_merge_scalar_mask):
Take elen instead of scalar BITS_PER_WORD.
(expand_vector_init_merge_repeating_sequence): Use inner_bits_size
instead of scaler BITS_PER_WORD.
---
gcc/config/riscv/riscv-v.cc | 10 ++++++----
1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/gcc/config/riscv/riscv-v.cc b/gcc/config/riscv/riscv-v.cc
index fb970344521..9270e258ca3 100644
--- a/gcc/config/riscv/riscv-v.cc
+++ b/gcc/config/riscv/riscv-v.cc
@@ -399,10 +399,11 @@ rvv_builder::get_merge_scalar_mask (unsigned int index_in_pattern) const
{
unsigned HOST_WIDE_INT mask = 0;
unsigned HOST_WIDE_INT base_mask = (1ULL << index_in_pattern);
+ unsigned int elen = TARGET_VECTOR_ELEN_64 ? 64 : 32;
- gcc_assert (BITS_PER_WORD % npatterns () == 0);
+ gcc_assert (elen % npatterns () == 0);
- int limit = BITS_PER_WORD / npatterns ();
+ int limit = elen / npatterns ();
for (int i = 0; i < limit; i++)
mask |= base_mask << (i * npatterns ());
@@ -1923,7 +1924,7 @@ expand_vector_init_merge_repeating_sequence (rtx target,
rtx mask = gen_reg_rtx (mask_mode);
rtx dup = gen_reg_rtx (dup_mode);
- if (full_nelts <= BITS_PER_WORD) /* vmv.s.x. */
+ if (full_nelts <= builder.inner_bits_size ()) /* vmv.s.x. */
{
rtx ops[] = {dup, gen_scalar_move_mask (dup_mask_mode),
RVV_VUNDEF (dup_mode), merge_mask};
@@ -1933,7 +1934,8 @@ expand_vector_init_merge_repeating_sequence (rtx target,
else /* vmv.v.x. */
{
rtx ops[] = {dup, force_reg (GET_MODE_INNER (dup_mode), merge_mask)};
- rtx vl = gen_int_mode (CEIL (full_nelts, BITS_PER_WORD), Pmode);
+ rtx vl = gen_int_mode (CEIL (full_nelts, builder.inner_bits_size ()),
+ Pmode);
emit_nonvlmax_integer_move_insn (code_for_pred_broadcast (dup_mode),
ops, vl);
}
--
2.34.1
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [PATCH v1] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32
2023-06-14 0:58 [PATCH v1] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32 pan2.li
@ 2023-06-14 1:07 ` juzhe.zhong
2023-06-14 7:30 ` Li, Pan2
2023-06-14 7:29 ` [PATCH v2] " pan2.li
2023-06-14 9:00 ` [PATCH v3] " pan2.li
2 siblings, 1 reply; 11+ messages in thread
From: juzhe.zhong @ 2023-06-14 1:07 UTC (permalink / raw)
To: pan2.li, gcc-patches
Cc: Robin Dapp, jeffreyalaw, pan2.li, yanzhang.wang, kito.cheng
[-- Attachment #1: Type: text/plain, Size: 2989 bytes --]
>> unsigned int elen = TARGET_VECTOR_ELEN_64 ? 64 : 32;
Add comment here to demonstrate why you pick up elen to set the LIMIT.
I understand:
1. -march=zve32* ===> ELEN = 32
-march=zve64* ===> ELEN = 64
2. both vmv.v.x/vmv.s.x is restrict to the ELEN
For example, When ELEN=32 (-march=zve32*)
vsetvli ...e64,m1
vmv.v.x/vmv.s.x
We can't support such code sequence.
You should demonstrate it clearly in the comments.
Otherwise, this patch LGTM.
juzhe.zhong@rivai.ai
From: pan2.li
Date: 2023-06-14 08:58
To: gcc-patches
CC: juzhe.zhong; rdapp.gcc; jeffreyalaw; pan2.li; yanzhang.wang; kito.cheng
Subject: [PATCH v1] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32
From: Pan Li <pan2.li@intel.com>
This patch would like to fix one bug exported by RV32 test case
multiple_rgroup_run-2.c. The mask should be restricted by elen in
vector, and the condition between the vmv.s.x and the vmv.v.x should
take inner_bits_size rather than constants.
Passed both the rv32 and rv64 riscv/rvv tests.
Signed-off-by: Pan Li <pan2.li@intel.com>
gcc/ChangeLog:
* config/riscv/riscv-v.cc (rvv_builder::get_merge_scalar_mask):
Take elen instead of scalar BITS_PER_WORD.
(expand_vector_init_merge_repeating_sequence): Use inner_bits_size
instead of scaler BITS_PER_WORD.
---
gcc/config/riscv/riscv-v.cc | 10 ++++++----
1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/gcc/config/riscv/riscv-v.cc b/gcc/config/riscv/riscv-v.cc
index fb970344521..9270e258ca3 100644
--- a/gcc/config/riscv/riscv-v.cc
+++ b/gcc/config/riscv/riscv-v.cc
@@ -399,10 +399,11 @@ rvv_builder::get_merge_scalar_mask (unsigned int index_in_pattern) const
{
unsigned HOST_WIDE_INT mask = 0;
unsigned HOST_WIDE_INT base_mask = (1ULL << index_in_pattern);
+ unsigned int elen = TARGET_VECTOR_ELEN_64 ? 64 : 32;
- gcc_assert (BITS_PER_WORD % npatterns () == 0);
+ gcc_assert (elen % npatterns () == 0);
- int limit = BITS_PER_WORD / npatterns ();
+ int limit = elen / npatterns ();
for (int i = 0; i < limit; i++)
mask |= base_mask << (i * npatterns ());
@@ -1923,7 +1924,7 @@ expand_vector_init_merge_repeating_sequence (rtx target,
rtx mask = gen_reg_rtx (mask_mode);
rtx dup = gen_reg_rtx (dup_mode);
- if (full_nelts <= BITS_PER_WORD) /* vmv.s.x. */
+ if (full_nelts <= builder.inner_bits_size ()) /* vmv.s.x. */
{
rtx ops[] = {dup, gen_scalar_move_mask (dup_mask_mode),
RVV_VUNDEF (dup_mode), merge_mask};
@@ -1933,7 +1934,8 @@ expand_vector_init_merge_repeating_sequence (rtx target,
else /* vmv.v.x. */
{
rtx ops[] = {dup, force_reg (GET_MODE_INNER (dup_mode), merge_mask)};
- rtx vl = gen_int_mode (CEIL (full_nelts, BITS_PER_WORD), Pmode);
+ rtx vl = gen_int_mode (CEIL (full_nelts, builder.inner_bits_size ()),
+ Pmode);
emit_nonvlmax_integer_move_insn (code_for_pred_broadcast (dup_mode),
ops, vl);
}
--
2.34.1
^ permalink raw reply [flat|nested] 11+ messages in thread
* [PATCH v2] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32
2023-06-14 0:58 [PATCH v1] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32 pan2.li
2023-06-14 1:07 ` juzhe.zhong
@ 2023-06-14 7:29 ` pan2.li
2023-06-14 7:43 ` juzhe.zhong
2023-06-14 8:27 ` Robin Dapp
2023-06-14 9:00 ` [PATCH v3] " pan2.li
2 siblings, 2 replies; 11+ messages in thread
From: pan2.li @ 2023-06-14 7:29 UTC (permalink / raw)
To: gcc-patches
Cc: juzhe.zhong, rdapp.gcc, jeffreyalaw, pan2.li, yanzhang.wang, kito.cheng
From: Pan Li <pan2.li@intel.com>
This patch would like to fix one bug exported by RV32 test case
multiple_rgroup_run-2.c. The mask should be restricted by elen in
vector, and the condition between the vmv.s.x and the vmv.v.x should
take inner_bits_size rather than constants.
After this patch, below failures on RV32 will be fixed.
FAIL: gcc.target/riscv/rvv/autovec/partial/multiple_rgroup_run-2.c execution test
FAIL: gcc.target/riscv/rvv/autovec/partial/multiple_rgroup_run-2.c execution test
FAIL: gcc.target/riscv/rvv/autovec/partial/multiple_rgroup_run-2.c execution test
FAIL: gcc.target/riscv/rvv/autovec/partial/multiple_rgroup_run-2.c execution test
FAIL: gcc.target/riscv/rvv/autovec/vls-vlmax/repeat_run-3.c -std=c99 -O3 -ftree-vectorize --param riscv-autovec-preference=fixed-vlmax execution test
Signed-off-by: Pan Li <pan2.li@intel.com>
gcc/ChangeLog:
* config/riscv/riscv-v.cc (rvv_builder::get_merge_scalar_mask):
Take elen instead of scalar BITS_PER_WORD.
(expand_vector_init_merge_repeating_sequence): Use inner_bits_size
instead of scaler BITS_PER_WORD.
---
gcc/config/riscv/riscv-v.cc | 18 ++++++++++++++----
1 file changed, 14 insertions(+), 4 deletions(-)
diff --git a/gcc/config/riscv/riscv-v.cc b/gcc/config/riscv/riscv-v.cc
index e07d5c2901a..db1a5529419 100644
--- a/gcc/config/riscv/riscv-v.cc
+++ b/gcc/config/riscv/riscv-v.cc
@@ -399,10 +399,19 @@ rvv_builder::get_merge_scalar_mask (unsigned int index_in_pattern) const
{
unsigned HOST_WIDE_INT mask = 0;
unsigned HOST_WIDE_INT base_mask = (1ULL << index_in_pattern);
+ /* We restrict the limit to the elen of RVV. For example:
+ -march=zve32*, the ELEN is 32.
+ -march=zve64*, the ELEN is 64.
+ The related vmv.v.x/vmv.s.x is restricted to ELEN as above, we cannot
+ take care of case like below when ELEN=32
+ vsetvil e64,m1
+ vmv.v.x/vmv.s.x
+ */
+ unsigned int elen = TARGET_VECTOR_ELEN_64 ? 64 : 32;
- gcc_assert (BITS_PER_WORD % npatterns () == 0);
+ gcc_assert (elen % npatterns () == 0);
- int limit = BITS_PER_WORD / npatterns ();
+ int limit = elen / npatterns ();
for (int i = 0; i < limit; i++)
mask |= base_mask << (i * npatterns ());
@@ -1928,7 +1937,7 @@ expand_vector_init_merge_repeating_sequence (rtx target,
rtx mask = gen_reg_rtx (mask_mode);
rtx dup = gen_reg_rtx (dup_mode);
- if (full_nelts <= BITS_PER_WORD) /* vmv.s.x. */
+ if (full_nelts <= builder.inner_bits_size ()) /* vmv.s.x. */
{
rtx ops[] = {dup, gen_scalar_move_mask (dup_mask_mode),
RVV_VUNDEF (dup_mode), merge_mask};
@@ -1938,7 +1947,8 @@ expand_vector_init_merge_repeating_sequence (rtx target,
else /* vmv.v.x. */
{
rtx ops[] = {dup, force_reg (GET_MODE_INNER (dup_mode), merge_mask)};
- rtx vl = gen_int_mode (CEIL (full_nelts, BITS_PER_WORD), Pmode);
+ rtx vl = gen_int_mode (CEIL (full_nelts, builder.inner_bits_size ()),
+ Pmode);
emit_nonvlmax_integer_move_insn (code_for_pred_broadcast (dup_mode),
ops, vl);
}
--
2.34.1
^ permalink raw reply [flat|nested] 11+ messages in thread
* RE: [PATCH v1] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32
2023-06-14 1:07 ` juzhe.zhong
@ 2023-06-14 7:30 ` Li, Pan2
0 siblings, 0 replies; 11+ messages in thread
From: Li, Pan2 @ 2023-06-14 7:30 UTC (permalink / raw)
To: juzhe.zhong, gcc-patches
Cc: Robin Dapp, jeffreyalaw, Wang, Yanzhang, kito.cheng
[-- Attachment #1: Type: text/plain, Size: 3897 bytes --]
Thanks Juzhe for reviewing, update the PATCH v2 as below.
https://gcc.gnu.org/pipermail/gcc-patches/2023-June/621645.html
Pan
From: juzhe.zhong@rivai.ai <juzhe.zhong@rivai.ai>
Sent: Wednesday, June 14, 2023 9:07 AM
To: Li, Pan2 <pan2.li@intel.com>; gcc-patches <gcc-patches@gcc.gnu.org>
Cc: Robin Dapp <rdapp.gcc@gmail.com>; jeffreyalaw <jeffreyalaw@gmail.com>; Li, Pan2 <pan2.li@intel.com>; Wang, Yanzhang <yanzhang.wang@intel.com>; kito.cheng <kito.cheng@gmail.com>
Subject: Re: [PATCH v1] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32
>> unsigned int elen = TARGET_VECTOR_ELEN_64 ? 64 : 32;
Add comment here to demonstrate why you pick up elen to set the LIMIT.
I understand:
1. -march=zve32* ===> ELEN = 32
-march=zve64* ===> ELEN = 64
2. both vmv.v.x/vmv.s.x is restrict to the ELEN
For example, When ELEN=32 (-march=zve32*)
vsetvli ...e64,m1
vmv.v.x/vmv.s.x
We can't support such code sequence.
You should demonstrate it clearly in the comments.
Otherwise, this patch LGTM.
________________________________
juzhe.zhong@rivai.ai<mailto:juzhe.zhong@rivai.ai>
From: pan2.li<mailto:pan2.li@intel.com>
Date: 2023-06-14 08:58
To: gcc-patches<mailto:gcc-patches@gcc.gnu.org>
CC: juzhe.zhong<mailto:juzhe.zhong@rivai.ai>; rdapp.gcc<mailto:rdapp.gcc@gmail.com>; jeffreyalaw<mailto:jeffreyalaw@gmail.com>; pan2.li<mailto:pan2.li@intel.com>; yanzhang.wang<mailto:yanzhang.wang@intel.com>; kito.cheng<mailto:kito.cheng@gmail.com>
Subject: [PATCH v1] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32
From: Pan Li <pan2.li@intel.com<mailto:pan2.li@intel.com>>
This patch would like to fix one bug exported by RV32 test case
multiple_rgroup_run-2.c. The mask should be restricted by elen in
vector, and the condition between the vmv.s.x and the vmv.v.x should
take inner_bits_size rather than constants.
Passed both the rv32 and rv64 riscv/rvv tests.
Signed-off-by: Pan Li <pan2.li@intel.com<mailto:pan2.li@intel.com>>
gcc/ChangeLog:
* config/riscv/riscv-v.cc (rvv_builder::get_merge_scalar_mask):
Take elen instead of scalar BITS_PER_WORD.
(expand_vector_init_merge_repeating_sequence): Use inner_bits_size
instead of scaler BITS_PER_WORD.
---
gcc/config/riscv/riscv-v.cc | 10 ++++++----
1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/gcc/config/riscv/riscv-v.cc b/gcc/config/riscv/riscv-v.cc
index fb970344521..9270e258ca3 100644
--- a/gcc/config/riscv/riscv-v.cc
+++ b/gcc/config/riscv/riscv-v.cc
@@ -399,10 +399,11 @@ rvv_builder::get_merge_scalar_mask (unsigned int index_in_pattern) const
{
unsigned HOST_WIDE_INT mask = 0;
unsigned HOST_WIDE_INT base_mask = (1ULL << index_in_pattern);
+ unsigned int elen = TARGET_VECTOR_ELEN_64 ? 64 : 32;
- gcc_assert (BITS_PER_WORD % npatterns () == 0);
+ gcc_assert (elen % npatterns () == 0);
- int limit = BITS_PER_WORD / npatterns ();
+ int limit = elen / npatterns ();
for (int i = 0; i < limit; i++)
mask |= base_mask << (i * npatterns ());
@@ -1923,7 +1924,7 @@ expand_vector_init_merge_repeating_sequence (rtx target,
rtx mask = gen_reg_rtx (mask_mode);
rtx dup = gen_reg_rtx (dup_mode);
- if (full_nelts <= BITS_PER_WORD) /* vmv.s.x. */
+ if (full_nelts <= builder.inner_bits_size ()) /* vmv.s.x. */
{
rtx ops[] = {dup, gen_scalar_move_mask (dup_mask_mode),
RVV_VUNDEF (dup_mode), merge_mask};
@@ -1933,7 +1934,8 @@ expand_vector_init_merge_repeating_sequence (rtx target,
else /* vmv.v.x. */
{
rtx ops[] = {dup, force_reg (GET_MODE_INNER (dup_mode), merge_mask)};
- rtx vl = gen_int_mode (CEIL (full_nelts, BITS_PER_WORD), Pmode);
+ rtx vl = gen_int_mode (CEIL (full_nelts, builder.inner_bits_size ()),
+ Pmode);
emit_nonvlmax_integer_move_insn (code_for_pred_broadcast (dup_mode),
ops, vl);
}
--
2.34.1
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [PATCH v2] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32
2023-06-14 7:29 ` [PATCH v2] " pan2.li
@ 2023-06-14 7:43 ` juzhe.zhong
2023-06-14 8:27 ` Robin Dapp
1 sibling, 0 replies; 11+ messages in thread
From: juzhe.zhong @ 2023-06-14 7:43 UTC (permalink / raw)
To: pan2.li, gcc-patches
Cc: Robin Dapp, jeffreyalaw, pan2.li, yanzhang.wang, kito.cheng
[-- Attachment #1: Type: text/plain, Size: 4011 bytes --]
+ /* We restrict the limit to the elen of RVV. For example:
+ -march=zve32*, the ELEN is 32.
+ -march=zve64*, the ELEN is 64.
+ The related vmv.v.x/vmv.s.x is restricted to ELEN as above, we cannot
+ take care of case like below when ELEN=32
+ vsetvil e64,m1
+ vmv.v.x/vmv.s.x
+ */
The comment is not clear enough.
How about:
According to RVV ISA SPEC, ELEN = 32 when -march=zve32* and ELEN = 64 when -march=zve64*.
Since vmv.v.x/vmv.s.x can't broadcast/move 64-bit value to the vector when ELEN = 32, we restrict the LIMIT to the ELEN.
I am not the native English speaker, I'd like to see Jeff or Robin comments that.
Thanks.
juzhe.zhong@rivai.ai
From: pan2.li
Date: 2023-06-14 15:29
To: gcc-patches
CC: juzhe.zhong; rdapp.gcc; jeffreyalaw; pan2.li; yanzhang.wang; kito.cheng
Subject: [PATCH v2] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32
From: Pan Li <pan2.li@intel.com>
This patch would like to fix one bug exported by RV32 test case
multiple_rgroup_run-2.c. The mask should be restricted by elen in
vector, and the condition between the vmv.s.x and the vmv.v.x should
take inner_bits_size rather than constants.
After this patch, below failures on RV32 will be fixed.
FAIL: gcc.target/riscv/rvv/autovec/partial/multiple_rgroup_run-2.c execution test
FAIL: gcc.target/riscv/rvv/autovec/partial/multiple_rgroup_run-2.c execution test
FAIL: gcc.target/riscv/rvv/autovec/partial/multiple_rgroup_run-2.c execution test
FAIL: gcc.target/riscv/rvv/autovec/partial/multiple_rgroup_run-2.c execution test
FAIL: gcc.target/riscv/rvv/autovec/vls-vlmax/repeat_run-3.c -std=c99 -O3 -ftree-vectorize --param riscv-autovec-preference=fixed-vlmax execution test
Signed-off-by: Pan Li <pan2.li@intel.com>
gcc/ChangeLog:
* config/riscv/riscv-v.cc (rvv_builder::get_merge_scalar_mask):
Take elen instead of scalar BITS_PER_WORD.
(expand_vector_init_merge_repeating_sequence): Use inner_bits_size
instead of scaler BITS_PER_WORD.
---
gcc/config/riscv/riscv-v.cc | 18 ++++++++++++++----
1 file changed, 14 insertions(+), 4 deletions(-)
diff --git a/gcc/config/riscv/riscv-v.cc b/gcc/config/riscv/riscv-v.cc
index e07d5c2901a..db1a5529419 100644
--- a/gcc/config/riscv/riscv-v.cc
+++ b/gcc/config/riscv/riscv-v.cc
@@ -399,10 +399,19 @@ rvv_builder::get_merge_scalar_mask (unsigned int index_in_pattern) const
{
unsigned HOST_WIDE_INT mask = 0;
unsigned HOST_WIDE_INT base_mask = (1ULL << index_in_pattern);
+ /* We restrict the limit to the elen of RVV. For example:
+ -march=zve32*, the ELEN is 32.
+ -march=zve64*, the ELEN is 64.
+ The related vmv.v.x/vmv.s.x is restricted to ELEN as above, we cannot
+ take care of case like below when ELEN=32
+ vsetvil e64,m1
+ vmv.v.x/vmv.s.x
+ */
+ unsigned int elen = TARGET_VECTOR_ELEN_64 ? 64 : 32;
- gcc_assert (BITS_PER_WORD % npatterns () == 0);
+ gcc_assert (elen % npatterns () == 0);
- int limit = BITS_PER_WORD / npatterns ();
+ int limit = elen / npatterns ();
for (int i = 0; i < limit; i++)
mask |= base_mask << (i * npatterns ());
@@ -1928,7 +1937,7 @@ expand_vector_init_merge_repeating_sequence (rtx target,
rtx mask = gen_reg_rtx (mask_mode);
rtx dup = gen_reg_rtx (dup_mode);
- if (full_nelts <= BITS_PER_WORD) /* vmv.s.x. */
+ if (full_nelts <= builder.inner_bits_size ()) /* vmv.s.x. */
{
rtx ops[] = {dup, gen_scalar_move_mask (dup_mask_mode),
RVV_VUNDEF (dup_mode), merge_mask};
@@ -1938,7 +1947,8 @@ expand_vector_init_merge_repeating_sequence (rtx target,
else /* vmv.v.x. */
{
rtx ops[] = {dup, force_reg (GET_MODE_INNER (dup_mode), merge_mask)};
- rtx vl = gen_int_mode (CEIL (full_nelts, BITS_PER_WORD), Pmode);
+ rtx vl = gen_int_mode (CEIL (full_nelts, builder.inner_bits_size ()),
+ Pmode);
emit_nonvlmax_integer_move_insn (code_for_pred_broadcast (dup_mode),
ops, vl);
}
--
2.34.1
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [PATCH v2] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32
2023-06-14 7:29 ` [PATCH v2] " pan2.li
2023-06-14 7:43 ` juzhe.zhong
@ 2023-06-14 8:27 ` Robin Dapp
2023-06-14 8:34 ` Li, Pan2
1 sibling, 1 reply; 11+ messages in thread
From: Robin Dapp @ 2023-06-14 8:27 UTC (permalink / raw)
To: pan2.li, gcc-patches
Cc: rdapp.gcc, juzhe.zhong, jeffreyalaw, yanzhang.wang, kito.cheng
Hi Pan,
> This patch would like to fix one bug exported by RV32 test case
> multiple_rgroup_run-2.c. The mask should be restricted by elen in
> vector, and the condition between the vmv.s.x and the vmv.v.x should
> take inner_bits_size rather than constants.
exported -> exposed.
How about something like:
"When constructing a vector mask from individual elements we wrongly
assumed that we can broadcast BITS_PER_WORD (i.e. XLEN). The maximum
is actually the vector element length (i.e. ELEN). This patch fixes
this."?
> + /* We restrict the limit to the elen of RVV. For example:
> + -march=zve32*, the ELEN is 32.
> + -march=zve64*, the ELEN is 64.
> + The related vmv.v.x/vmv.s.x is restricted to ELEN as above, we cannot
> + take care of case like below when ELEN=32
> + vsetvil e64,m1
> + vmv.v.x/vmv.s.x
> + */
/* Here we construct a mask pattern that will later be broadcast
to a vector register. The maximum broadcast size for vmv.v.x/vmv.s.x
is determined by the length of a vector element (ELEN) and not by
XLEN so make sure we do not exceed it. One example is -march=zve32*
which mandates ELEN == 32 but can be combined with -march=rv64
with XLEN == 64. */
Regards
Robin
^ permalink raw reply [flat|nested] 11+ messages in thread
* RE: [PATCH v2] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32
2023-06-14 8:27 ` Robin Dapp
@ 2023-06-14 8:34 ` Li, Pan2
0 siblings, 0 replies; 11+ messages in thread
From: Li, Pan2 @ 2023-06-14 8:34 UTC (permalink / raw)
To: Robin Dapp, gcc-patches
Cc: juzhe.zhong, jeffreyalaw, Wang, Yanzhang, kito.cheng
Thanks Robin, that looks like much better than the v2, let me update it to PATCH v3.
Pan
-----Original Message-----
From: Robin Dapp <rdapp.gcc@gmail.com>
Sent: Wednesday, June 14, 2023 4:27 PM
To: Li, Pan2 <pan2.li@intel.com>; gcc-patches@gcc.gnu.org
Cc: rdapp.gcc@gmail.com; juzhe.zhong@rivai.ai; jeffreyalaw@gmail.com; Wang, Yanzhang <yanzhang.wang@intel.com>; kito.cheng@gmail.com
Subject: Re: [PATCH v2] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32
Hi Pan,
> This patch would like to fix one bug exported by RV32 test case
> multiple_rgroup_run-2.c. The mask should be restricted by elen in
> vector, and the condition between the vmv.s.x and the vmv.v.x should
> take inner_bits_size rather than constants.
exported -> exposed.
How about something like:
"When constructing a vector mask from individual elements we wrongly
assumed that we can broadcast BITS_PER_WORD (i.e. XLEN). The maximum
is actually the vector element length (i.e. ELEN). This patch fixes
this."?
> + /* We restrict the limit to the elen of RVV. For example:
> + -march=zve32*, the ELEN is 32.
> + -march=zve64*, the ELEN is 64.
> + The related vmv.v.x/vmv.s.x is restricted to ELEN as above, we cannot
> + take care of case like below when ELEN=32
> + vsetvil e64,m1
> + vmv.v.x/vmv.s.x
> + */
/* Here we construct a mask pattern that will later be broadcast
to a vector register. The maximum broadcast size for vmv.v.x/vmv.s.x
is determined by the length of a vector element (ELEN) and not by
XLEN so make sure we do not exceed it. One example is -march=zve32*
which mandates ELEN == 32 but can be combined with -march=rv64
with XLEN == 64. */
Regards
Robin
^ permalink raw reply [flat|nested] 11+ messages in thread
* [PATCH v3] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32
2023-06-14 0:58 [PATCH v1] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32 pan2.li
2023-06-14 1:07 ` juzhe.zhong
2023-06-14 7:29 ` [PATCH v2] " pan2.li
@ 2023-06-14 9:00 ` pan2.li
2023-06-14 9:01 ` juzhe.zhong
2 siblings, 1 reply; 11+ messages in thread
From: pan2.li @ 2023-06-14 9:00 UTC (permalink / raw)
To: gcc-patches
Cc: juzhe.zhong, rdapp.gcc, jeffreyalaw, pan2.li, yanzhang.wang, kito.cheng
From: Pan Li <pan2.li@intel.com>
When constructing a vector mask from individual elements we wrongly
assumed that we can broadcast BITS_PER_WORD (i.e. XLEN). The maximum is
actually the vector element length (i.e. ELEN). This patch fixes this.
After this patch, below failures on RV32 will be fixed.
FAIL: gcc.target/riscv/rvv/autovec/partial/multiple_rgroup_run-2.c execution test
FAIL: gcc.target/riscv/rvv/autovec/partial/multiple_rgroup_run-2.c execution test
FAIL: gcc.target/riscv/rvv/autovec/partial/multiple_rgroup_run-2.c execution test
FAIL: gcc.target/riscv/rvv/autovec/partial/multiple_rgroup_run-2.c execution test
FAIL: gcc.target/riscv/rvv/autovec/vls-vlmax/repeat_run-3.c -std=c99 -O3 -ftree-vectorize --param riscv-autovec-preference=fixed-vlmax execution test
Signed-off-by: Pan Li <pan2.li@intel.com>
gcc/ChangeLog:
* config/riscv/riscv-v.cc (rvv_builder::get_merge_scalar_mask):
Take elen instead of scalar BITS_PER_WORD.
(expand_vector_init_merge_repeating_sequence): Use inner_bits_size
instead of scaler BITS_PER_WORD.
---
gcc/config/riscv/riscv-v.cc | 16 ++++++++++++----
1 file changed, 12 insertions(+), 4 deletions(-)
diff --git a/gcc/config/riscv/riscv-v.cc b/gcc/config/riscv/riscv-v.cc
index e07d5c2901a..01f647bc0bd 100644
--- a/gcc/config/riscv/riscv-v.cc
+++ b/gcc/config/riscv/riscv-v.cc
@@ -399,10 +399,17 @@ rvv_builder::get_merge_scalar_mask (unsigned int index_in_pattern) const
{
unsigned HOST_WIDE_INT mask = 0;
unsigned HOST_WIDE_INT base_mask = (1ULL << index_in_pattern);
+ /* Here we construct a mask pattern that will later be broadcast
+ to a vector register. The maximum broadcast size for vmv.v.x/vmv.s.x
+ is determined by the length of a vector element (ELEN) and not by
+ XLEN so make sure we do not exceed it. One example is -march=zve32*
+ which mandates ELEN == 32 but can be combined with -march=rv64
+ with XLEN == 64. */
+ unsigned int elen = TARGET_VECTOR_ELEN_64 ? 64 : 32;
- gcc_assert (BITS_PER_WORD % npatterns () == 0);
+ gcc_assert (elen % npatterns () == 0);
- int limit = BITS_PER_WORD / npatterns ();
+ int limit = elen / npatterns ();
for (int i = 0; i < limit; i++)
mask |= base_mask << (i * npatterns ());
@@ -1928,7 +1935,7 @@ expand_vector_init_merge_repeating_sequence (rtx target,
rtx mask = gen_reg_rtx (mask_mode);
rtx dup = gen_reg_rtx (dup_mode);
- if (full_nelts <= BITS_PER_WORD) /* vmv.s.x. */
+ if (full_nelts <= builder.inner_bits_size ()) /* vmv.s.x. */
{
rtx ops[] = {dup, gen_scalar_move_mask (dup_mask_mode),
RVV_VUNDEF (dup_mode), merge_mask};
@@ -1938,7 +1945,8 @@ expand_vector_init_merge_repeating_sequence (rtx target,
else /* vmv.v.x. */
{
rtx ops[] = {dup, force_reg (GET_MODE_INNER (dup_mode), merge_mask)};
- rtx vl = gen_int_mode (CEIL (full_nelts, BITS_PER_WORD), Pmode);
+ rtx vl = gen_int_mode (CEIL (full_nelts, builder.inner_bits_size ()),
+ Pmode);
emit_nonvlmax_integer_move_insn (code_for_pred_broadcast (dup_mode),
ops, vl);
}
--
2.34.1
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [PATCH v3] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32
2023-06-14 9:00 ` [PATCH v3] " pan2.li
@ 2023-06-14 9:01 ` juzhe.zhong
2023-06-14 18:56 ` Jeff Law
0 siblings, 1 reply; 11+ messages in thread
From: juzhe.zhong @ 2023-06-14 9:01 UTC (permalink / raw)
To: pan2.li, gcc-patches
Cc: Robin Dapp, jeffreyalaw, pan2.li, yanzhang.wang, kito.cheng
[-- Attachment #1: Type: text/plain, Size: 3383 bytes --]
LGTM
juzhe.zhong@rivai.ai
From: pan2.li
Date: 2023-06-14 17:00
To: gcc-patches
CC: juzhe.zhong; rdapp.gcc; jeffreyalaw; pan2.li; yanzhang.wang; kito.cheng
Subject: [PATCH v3] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32
From: Pan Li <pan2.li@intel.com>
When constructing a vector mask from individual elements we wrongly
assumed that we can broadcast BITS_PER_WORD (i.e. XLEN). The maximum is
actually the vector element length (i.e. ELEN). This patch fixes this.
After this patch, below failures on RV32 will be fixed.
FAIL: gcc.target/riscv/rvv/autovec/partial/multiple_rgroup_run-2.c execution test
FAIL: gcc.target/riscv/rvv/autovec/partial/multiple_rgroup_run-2.c execution test
FAIL: gcc.target/riscv/rvv/autovec/partial/multiple_rgroup_run-2.c execution test
FAIL: gcc.target/riscv/rvv/autovec/partial/multiple_rgroup_run-2.c execution test
FAIL: gcc.target/riscv/rvv/autovec/vls-vlmax/repeat_run-3.c -std=c99 -O3 -ftree-vectorize --param riscv-autovec-preference=fixed-vlmax execution test
Signed-off-by: Pan Li <pan2.li@intel.com>
gcc/ChangeLog:
* config/riscv/riscv-v.cc (rvv_builder::get_merge_scalar_mask):
Take elen instead of scalar BITS_PER_WORD.
(expand_vector_init_merge_repeating_sequence): Use inner_bits_size
instead of scaler BITS_PER_WORD.
---
gcc/config/riscv/riscv-v.cc | 16 ++++++++++++----
1 file changed, 12 insertions(+), 4 deletions(-)
diff --git a/gcc/config/riscv/riscv-v.cc b/gcc/config/riscv/riscv-v.cc
index e07d5c2901a..01f647bc0bd 100644
--- a/gcc/config/riscv/riscv-v.cc
+++ b/gcc/config/riscv/riscv-v.cc
@@ -399,10 +399,17 @@ rvv_builder::get_merge_scalar_mask (unsigned int index_in_pattern) const
{
unsigned HOST_WIDE_INT mask = 0;
unsigned HOST_WIDE_INT base_mask = (1ULL << index_in_pattern);
+ /* Here we construct a mask pattern that will later be broadcast
+ to a vector register. The maximum broadcast size for vmv.v.x/vmv.s.x
+ is determined by the length of a vector element (ELEN) and not by
+ XLEN so make sure we do not exceed it. One example is -march=zve32*
+ which mandates ELEN == 32 but can be combined with -march=rv64
+ with XLEN == 64. */
+ unsigned int elen = TARGET_VECTOR_ELEN_64 ? 64 : 32;
- gcc_assert (BITS_PER_WORD % npatterns () == 0);
+ gcc_assert (elen % npatterns () == 0);
- int limit = BITS_PER_WORD / npatterns ();
+ int limit = elen / npatterns ();
for (int i = 0; i < limit; i++)
mask |= base_mask << (i * npatterns ());
@@ -1928,7 +1935,7 @@ expand_vector_init_merge_repeating_sequence (rtx target,
rtx mask = gen_reg_rtx (mask_mode);
rtx dup = gen_reg_rtx (dup_mode);
- if (full_nelts <= BITS_PER_WORD) /* vmv.s.x. */
+ if (full_nelts <= builder.inner_bits_size ()) /* vmv.s.x. */
{
rtx ops[] = {dup, gen_scalar_move_mask (dup_mask_mode),
RVV_VUNDEF (dup_mode), merge_mask};
@@ -1938,7 +1945,8 @@ expand_vector_init_merge_repeating_sequence (rtx target,
else /* vmv.v.x. */
{
rtx ops[] = {dup, force_reg (GET_MODE_INNER (dup_mode), merge_mask)};
- rtx vl = gen_int_mode (CEIL (full_nelts, BITS_PER_WORD), Pmode);
+ rtx vl = gen_int_mode (CEIL (full_nelts, builder.inner_bits_size ()),
+ Pmode);
emit_nonvlmax_integer_move_insn (code_for_pred_broadcast (dup_mode),
ops, vl);
}
--
2.34.1
^ permalink raw reply [flat|nested] 11+ messages in thread
* Re: [PATCH v3] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32
2023-06-14 9:01 ` juzhe.zhong
@ 2023-06-14 18:56 ` Jeff Law
2023-06-15 1:05 ` Li, Pan2
0 siblings, 1 reply; 11+ messages in thread
From: Jeff Law @ 2023-06-14 18:56 UTC (permalink / raw)
To: juzhe.zhong, pan2.li, gcc-patches; +Cc: Robin Dapp, yanzhang.wang, kito.cheng
On 6/14/23 03:01, juzhe.zhong@rivai.ai wrote:
> LGTM
Agreed. Commit when convenient.
jeff
^ permalink raw reply [flat|nested] 11+ messages in thread
* RE: [PATCH v3] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32
2023-06-14 18:56 ` Jeff Law
@ 2023-06-15 1:05 ` Li, Pan2
0 siblings, 0 replies; 11+ messages in thread
From: Li, Pan2 @ 2023-06-15 1:05 UTC (permalink / raw)
To: Jeff Law, juzhe.zhong, gcc-patches; +Cc: Robin Dapp, Wang, Yanzhang, kito.cheng
Committed, thanks Jeff and Juzhe.
Pan
-----Original Message-----
From: Jeff Law <jeffreyalaw@gmail.com>
Sent: Thursday, June 15, 2023 2:56 AM
To: juzhe.zhong@rivai.ai; Li, Pan2 <pan2.li@intel.com>; gcc-patches <gcc-patches@gcc.gnu.org>
Cc: Robin Dapp <rdapp.gcc@gmail.com>; Wang, Yanzhang <yanzhang.wang@intel.com>; kito.cheng <kito.cheng@gmail.com>
Subject: Re: [PATCH v3] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32
On 6/14/23 03:01, juzhe.zhong@rivai.ai wrote:
> LGTM
Agreed. Commit when convenient.
jeff
^ permalink raw reply [flat|nested] 11+ messages in thread
end of thread, other threads:[~2023-06-15 1:06 UTC | newest]
Thread overview: 11+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-06-14 0:58 [PATCH v1] RISC-V: Bugfix for vec_init repeating auto vectorization in RV32 pan2.li
2023-06-14 1:07 ` juzhe.zhong
2023-06-14 7:30 ` Li, Pan2
2023-06-14 7:29 ` [PATCH v2] " pan2.li
2023-06-14 7:43 ` juzhe.zhong
2023-06-14 8:27 ` Robin Dapp
2023-06-14 8:34 ` Li, Pan2
2023-06-14 9:00 ` [PATCH v3] " pan2.li
2023-06-14 9:01 ` juzhe.zhong
2023-06-14 18:56 ` Jeff Law
2023-06-15 1:05 ` Li, Pan2
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).