Earlier, we were gating SVE2 faminmax behind sve+faminmax. This was
incorrect and this patch changes it so that it is gated behind
sve2+faminmax.
gcc/ChangeLog:
* config/aarch64/aarch64-sve2.md:
(*aarch64_pred_faminmax_fused): Fix to use the correct flags.
* config/aarch6
This patch introduces support for LUTI2/LUTI4 ACLE for SVE2.
LUTI instructions are used for efficient table lookups with 2-bit
or 4-bit indices. LUTI2 reads indexed 8-bit or 16-bit elements from
the low 128 bits of the table vector using packed 2-bit indices,
while LUTI4 can read from the low 128
This patch introduces support for LUTI2/LUTI4 ACLE for SVE2.
LUTI instructions are used for efficient table lookups with 2-bit
or 4-bit indices. LUTI2 reads indexed 8-bit or 16-bit elements from
the low 128 bits of the table vector using packed 2-bit indices,
while LUTI4 can read from the low 128
This patch introduces support for LUTI2/LUTI4 ACLE for SVE2.
LUTI instructions are used for efficient table lookups with 2-bit
or 4-bit indices. LUTI2 reads indexed 8-bit or 16-bit elements from
the low 128 bits of the table vector using packed 2-bit indices,
while LUTI4 can read from the low 128
This patch introduces support for LUTI2/LUTI4 ACLE for SVE2.
LUTI instructions are used for efficient table lookups with 2-bit
or 4-bit indices. LUTI2 reads indexed 8-bit or 16-bit elements from
the low 128 bits of the table vector using packed 2-bit indices,
while LUTI4 can read from the low 128
The AArch64 FEAT_LUT extension is optional from Armv9.2-a and mandatory
from Armv9.5-a. It introduces instructions for lookup table reads with
bit indices.
This patch adds support for AdvSIMD lut intrinsics. The intrinsics for
this extension are implemented as the following builtin functions:
* v
Refactor AdvSIMD intrinsics defined using the new pragma-based approach
so that it is more extensible.
Introduce a new struct, simd_type, which defines types using a mode and
qualifiers, and use objects of this struct in the declaration of intrinsics
in the aarch64-simd-pragma-builtins.def file.
From: Saurabh Jha
This patch series is a revised version of:
https://gcc.gnu.org/pipermail/gcc-patches/2024-November/667692.html
In the refactor patch, I redesigned how types are declared and how
expand happens. I have taken some ideas from the reviews of the
other patch series:
https://gcc.gnu.
The AArch64 FEAT_FP8DOT2 and FEAT_FP8DOT4 extension introduces
instructions for dot product of vectors.
This patch introduces the following intrinsics:
1. vdot{q}_{fp16|fp32}_mf8_fpm.
2. vdot{q}_lane{q}_{fp16|fp32}_mf8_fpm.
It introduces two flags: fp8dot2 and fp8dot4.
We had to add space for a
The AArch64 FEAT_FP8 extension introduces instructions for conversion
and scaling.
This patch introduces the following intrinsics:
1. vcvt{1|2}_{bf16|high_bf16|low_bf16}_mf8_fpm.
2. vcvt{q}_mf8_f16_fpm.
3. vcvt_{high}_mf8_f32_fpm.
4. vscale{q}_{f16|f32|f64}.
We introduced two aarch64_builtin_sig
The AArch64 FEAT_FP8FMA extension introduces instructions for
multiply-add of vectors.
This patch introduces the following instructions:
1. {vmlalbq|vmlaltq}_f16_mf8_fpm.
2. {vmlalbq|vmlaltq}_lane{q}_f16_mf8_fpm.
3. {vmlallbbq|vmlallbtq|vmlalltbq|vmlallttq}_f32_mf8_fpm.
4. {vmlallbbq|vmlallbtq|vm
From: Saurabh Jha
This patch series is a revised version of:
https://gcc.gnu.org/pipermail/gcc-patches/2024-November/667723.html
I have addressed comments around building a list of operands while
declaring types and while expanding to RTL. I have also removed
signatures with "_fpm" and "_lane" a
The AArch64 FEAT_FP8FMA extension introduces instructions for
multiply-add of vectors.
This patch introduces the following instructions:
1. {vmlalbq|vmlaltq}_f16_mf8_fpm.
2. {vmlalbq|vmlaltq}_lane{q}_f16_mf8_fpm.
3. {vmlallbbq|vmlallbtq|vmlalltbq|vmlallttq}_f32_mf8_fpm.
4. {vmlallbbq|vmlallbtq|vm
The AArch64 FEAT_FP8DOT2 and FEAT_FP8DOT4 extension introduces
instructions for dot product of vectors.
This patch introduces the following intrinsics:
1. vdot{q}_{fp16|fp32}_mf8_fpm.
2. vdot{q}_lane{q}_{fp16|fp32}_mf8_fpm.
It introduces two flags: fp8dot2 and fp8dot4.
We had to add space for a
The AArch64 FEAT_FP8 extension introduces instructions for conversion
and scaling.
This patch introduces the following intrinsics:
1. vcvt{1|2}_{bf16|high_bf16|low_bf16}_mf8_fpm.
2. vcvt{q}_mf8_f16_fpm.
3. vcvt_{high}_mf8_f32_fpm.
4. vscale{q}_{f16|f32|f64}.
We introduced three new aarch64_built
From: Saurabh Jha
This patch series has three patches for adding support for fp8, fp8dot2,
and fp8dot4 acle AdvSIMD intrinsics.
The specific things I need thoughts are on are written in each commit
message after the "---" which will be omitted when commiting.
Regression tested on aarch64-unknow
The AArch64 FEAT_FAMINMAX extension introduces instructions for
computing the floating point absolute maximum and minimum of the
two vectors element-wise.
This patch introduces SVE2 faminmax intrinsics. The intrinsics of this
extension are implemented as the following builtin functions:
* sva[max
The AArch64 FEAT_FAMINMAX extension introduces instructions for
computing the floating point absolute maximum and minimum of the
two vectors element-wise.
This patch adds code generation for famax and famin in terms of existing
unspecs. With this patch:
1. famax can be expressed as taking UNSPEC_
From: Saurabh Jha
This patch series is a revised version of:
https://gcc.gnu.org/pipermail/gcc-patches/2024-October/664912.html
In particular, the only changes are in the first patch, where in the
test cases of intrinsics, we removed unnecessary capture of regular
expression of operands. The sec
The AArch64 FEAT_FAMINMAX extension introduces instructions for
computing the floating point absolute maximum and minimum of the
two vectors element-wise.
This patch adds code generation for famax and famin in terms of existing
unspecs. With this patch:
1. famax can be expressed as taking UNSPEC_
From: Saurabh Jha
This patch series is a revised version of:
https://gcc.gnu.org/pipermail/gcc-patches/2024-October/664391.html
Previous review comments are in this thread:
https://gcc.gnu.org/pipermail/gcc-patches/2024-October/664329.html
The second patch of this is okay to merge. The changes
The AArch64 FEAT_FAMINMAX extension introduces instructions for
computing the floating point absolute maximum and minimum of the
two vectors element-wise.
This patch introduces SVE2 faminmax intrinsics. The intrinsics of this
extension are implemented as the following builtin functions:
* sva[max
In ac4cdf5cb43c0b09e81760e2a1902ceebcf1a135, I introduced a bug where
I put the new unspecs, UNSPEC_COND_SMAX and UNSPEC_COND_SMIN, into the
wrong iterator.
I should have put new unspecs in SVE_COND_FP_MAXMIN but I put it in
SVE_COND_FP_BINARY_REG instead. That was incorrect because the
SVE_COND_
The AArch64 FEAT_FAMINMAX extension introduces instructions for
computing the floating point absolute maximum and minimum of the
two vectors element-wise.
This patch introduces SVE2 faminmax intrinsics. The intrinsics of this
extension are implemented as the following builtin functions:
* sva[max
The AArch64 FEAT_FAMINMAX extension introduces instructions for
computing the floating point absolute maximum and minimum of the
two vectors element-wise.
This patch adds code generation for famax and famin in terms of existing
unspecs. With this patch:
1. famax can be expressed as taking UNSPEC_
From: Saurabh Jha
This is a revised version of this patch series:
https://gcc.gnu.org/pipermail/gcc-patches/2024-October/664329.html
Unfortunately, I had test case failures which I missed but shouldn't
have. Apologies for that.
This version fixes the failing test cases in the second patch with
The AArch64 FEAT_FAMINMAX extension introduces instructions for
computing the floating point absolute maximum and minimum of the
two vectors element-wise.
This patch introduces SVE2 faminmax intrinsics. The intrinsics of this
extension are implemented as the following builtin functions:
* sva[max
The AArch64 FEAT_FAMINMAX extension introduces instructions for
computing the floating point absolute maximum and minimum of the
two vectors element-wise.
This patch adds code generation for famax and famin in terms of existing
unspecs. With this patch:
1. famax can be expressed as taking UNSPEC_
From: Saurabh Jha
This patch series is a revised version of:
https://gcc.gnu.org/pipermail/gcc-patches/2024-October/664209.html
The second commit of the previous patch series was reviewed and has been
commited separately. This patch contains first and third commit of the
previous patch.
The cha
Introduce two new unspecs, UNSPEC_COND_SMAX and UNSPEC_COND_SMIN,
corresponding to rtl operators smax and smin. UNSPEC_COND_SMAX is used
to generate fmaxnm instruction and UNSPEC_COND_SMIN is used to generate
fminnm instruction.
With these new unspecs, we can generate SVE2 max/min instructions us
The AArch64 FEAT_FAMINMAX extension introduces instructions for
computing the floating point absolute maximum and minimum of the
two vectors element-wise.
This patch adds code generation for famax and famin in terms of existing
unspecs. With this patch:
1. famax can be expressed as taking UNSPEC_
The AArch64 FEAT_FAMINMAX extension introduces instructions for
computing the floating point absolute maximum and minimum of the
two vectors element-wise.
This patch introduces SVE2 faminmax intrinsics. The intrinsics of this
extension are implemented as the following builtin functions:
* sva[max
From: Saurabh Jha
This patch series a revised version of an earlier patch series:
https://gcc.gnu.org/pipermail/gcc-patches/2024-September/662951.html.
The main change in this patch series is the introduction of the two
unspecs, UNSPEC_COND_SMAX and UNSPEC_COND_SMIN, and using them for
existing
ChangeLog:
* MAINTAINERS: Fix sort order and add username.
---
MAINTAINERS | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/MAINTAINERS b/MAINTAINERS
index 0ea4db20f88..3b4cf9d20d8 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -554,10 +554,10 @@ Sam James
ChangeLog:
* MAINTAINERS: Add myself to write after approval.
---
MAINTAINERS | 1 +
1 file changed, 1 insertion(+)
diff --git a/MAINTAINERS b/MAINTAINERS
index e9fafaf45a7..0ea4db20f88 100644
--- a/MAINTAINERS
+++ b/MAINTAINERS
@@ -557,6 +557,7 @@ Andrew Jenner andrew
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors
element-wise.
This patch adds code generation support for famax and famin in terms of
existing R
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors element-wise.
This patch introduces AdvSIMD faminmax intrinsics. The intrinsics of
this extensio
From: Saurabh Jha
This is a revised version of this patch series:
https://gcc.gnu.org/pipermail/gcc-patches/2024-September/663204.html
The only new thing in both patches of this series are fixing directives
in test cases, replace /* { dg-do assemble} */ with
/* { dg-do compile } */. We need comp
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors
element-wise.
This patch adds code generation support for famax and famin in terms of
existing R
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors element-wise.
This patch introduces AdvSIMD faminmax intrinsics. The intrinsics of
this extensio
From: Saurabh Jha
This is a revised version of this patch series:
https://gcc.gnu.org/pipermail/gcc-patches/2024-September/thread.html
Thanks for the review comments. They are all addressed in this version.
The changes are as follows.
1. [intrinsics patch] Using enum class for aarch64_builtin_si
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors
element-wise.
This patch adds code generation for famax and famin in terms of existing
unspecs.
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors
element-wise.
This patch introduces SVE2 faminmax intrinsics. The intrinsics of this
extension a
From: Saurabh Jha
This patch series adds support for SVE2 faminmax. It should be merged
only after AdvSIMD faminmax patch series is merged:
https://gcc.gnu.org/pipermail/gcc-patches/2024-September/662131.html
The first patch adds intrinsics and the second patch adds support for
combining FMAX/FM
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors
element-wise.
This patch adds code generation support for famax and famin in terms of
existing R
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors element-wise.
This patch introduces AdvSIMD faminmax intrinsics. The intrinsics of
this extensio
From: Saurabh Jha
This series is a revised version of:
https://gcc.gnu.org/pipermail/gcc-patches/2024-August/661860.html.
The first patch of the series is updated to address these comments:
https://gcc.gnu.org/pipermail/gcc-patches/2024-August/661866.html
All comments are addressed exactly as s
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors
element-wise.
This patch adds code generation support for famax and famin in terms of
existing R
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors element-wise.
This patch introduces AdvSIMD faminmax intrinsics. The intrinsics of
this extensio
From: Saurabh Jha
This patch series is a respin of
https://gcc.gnu.org/pipermail/gcc-patches/2024-August/661757.html.
The major refactorings suggested in the reviews to previous version will
be done separately to keep the scope of this series small. I'll create a
new series for that refactoring.
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors
element-wise.
This patch adds code generation support for famax and famin in terms of
existing R
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors element-wise.
This patch introduces AdvSIMD faminmax intrinsics. The intrinsics of
this extensio
From: Saurabh Jha
This patch series is a respin of
https://gcc.gnu.org/pipermail/gcc-patches/2024-August/661672.html.
The new version addresses comment about using AARCH64_PRAGMA_BUILTIN_START
and AARCH64_PRAGMA_BUILTIN_END in aarch64_builtins enum. Apart from the
function expand_pragma_builtin
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors element-wise.
This patch introduces AdvSIMD faminmax intrinsics. The intrinsics of
this extensio
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors
element-wise.
This patch adds code generation support for famax and famin in terms of
existing R
From: Saurabh Jha
This patch series is a respin of the previous patch here:
https://gcc.gnu.org/pipermail/gcc-patches/2024-August/660917.html
The new version addresses review comments on the previous patch series.
It also introduced a new way of defining AArch4 AdvSIMD intrinsics. All
of the new
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors
element-wise.
This patch adds code generation support for famax and famin in terms of
existing R
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors element-wise.
This patch introduces AdvSIMD faminmax intrinsics. The intrinsics of
this extensio
From: Saurabh Jha
This patch series is a respin of the previous patch here:
https://gcc.gnu.org/pipermail/gcc-patches/2024-August/659749.html.
This new version is rebased with latest master after the merging
of this patch series:
https://gcc.gnu.org/pipermail/gcc-patches/2024-August/660532.html.
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors element-wise.
This patch does two things:
1. Introduces AdvSIMD faminmax intrinsics.
2. Move rep
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors
element-wise.
This patch adds code generation support for famax and famin in terms of
existing R
From: Saurabh Jha
This patch series is a respin of a previous patch here:
https://gcc.gnu.org/pipermail/gcc-patches/2024-August/658984.html
The AArch64 FEAT_FAMINMAX is optional from Armv9.2-a and mandatory from
Armv9.5-a. It introduces instructions for computing the floating point
absolute maxi
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors element-wise.
This patch does three things:
1. Introduces AdvSIMD faminmax intrinsics.
2. Adds c
The AArch64 FEAT_FAMINMAX extension is optional from Armv9.2-a and
mandatory from Armv9.5-a. It introduces instructions for computing the
floating point absolute maximum and minimum of the two vectors element-wise.
This patch does three things:
1. Introduces AdvSIMD faminmax intrinsics.
2. Adds c
The AArch64 FEAT_FAMINMAX extension is optional in Armv9.2 and mandatory
in Armv9.5. It introduces instructions for computing the floating point
absolute maximum and minimum of the two vectors element-wise.
This patch introduces intrinsics for AdvSIMD faminmax extension in the
form of the followi
65 matches
Mail list logo
