This patch is to update {skylake,icelake,alderlake}_cost to add a bit
preference to vector store.
Since the interger vector construction cost has changed, we need to adjust the
load and store costs for intel processers.
With the patch applied
538.imagic_r:gets ~6% improvement on ADL for
From: zhongjuzhe
gcc/testsuite/ChangeLog:
* g++.target/riscv/rvv/vssex.C: New test.
---
gcc/testsuite/g++.target/riscv/rvv/vssex.C | 1704
1 file changed, 1704 insertions(+)
create mode 100644 gcc/testsuite/g++.target/riscv/rvv/vssex.C
diff --git
From: zhongjuzhe
gcc/testsuite/ChangeLog:
* g++.target/riscv/rvv/vluxeix_4.C: New test.
---
.../g++.target/riscv/rvv/vluxeix_4.C | 2503 +
1 file changed, 2503 insertions(+)
create mode 100644 gcc/testsuite/g++.target/riscv/rvv/vluxeix_4.C
diff --git
From: zhongjuzhe
gcc/testsuite/ChangeLog:
* g++.target/riscv/rvv/vsex.C: New test.
---
gcc/testsuite/g++.target/riscv/rvv/vsex.C | 1704 +
1 file changed, 1704 insertions(+)
create mode 100644 gcc/testsuite/g++.target/riscv/rvv/vsex.C
diff --git
From: zhongjuzhe
gcc/testsuite/ChangeLog:
* g++.target/riscv/rvv/vloxeix_4.C: New test.
---
.../g++.target/riscv/rvv/vloxeix_4.C | 2503 +
1 file changed, 2503 insertions(+)
create mode 100644 gcc/testsuite/g++.target/riscv/rvv/vloxeix_4.C
diff --git
From: zhongjuzhe
gcc/testsuite/ChangeLog:
* gcc.target/riscv/rvv/intrinsic/vsex.c: New test.
---
.../gcc.target/riscv/rvv/intrinsic/vsex.c | 4776 +
1 file changed, 4776 insertions(+)
create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/intrinsic/vsex.c
diff
From: zhongjuzhe
gcc/testsuite/ChangeLog:
* gcc.target/riscv/rvv/intrinsic/vlsex_2.c: New test.
---
.../gcc.target/riscv/rvv/intrinsic/vlsex_2.c | 1251 +
1 file changed, 1251 insertions(+)
create mode 100644
From: zhongjuzhe
gcc/testsuite/ChangeLog:
* gcc.target/riscv/rvv/intrinsic/vlexff_2.c: New test.
---
.../gcc.target/riscv/rvv/intrinsic/vlexff_2.c | 1251 +
1 file changed, 1251 insertions(+)
create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/intrinsic/vlexff_2.c
From: zhongjuzhe
gcc/testsuite/ChangeLog:
* gcc.target/riscv/rvv/intrinsic/mask_load_store.c: New test.
* gcc.target/riscv/rvv/intrinsic/mask_load_store_31.c: New test.
* gcc.target/riscv/rvv/intrinsic/mask_load_store_32.c: New test.
---
From: zhongjuzhe
gcc/testsuite/ChangeLog:
* gcc.target/riscv/rvv/intrinsic/vlex_2.c: New test.
---
.../gcc.target/riscv/rvv/intrinsic/vlex_2.c | 1251 +
1 file changed, 1251 insertions(+)
create mode 100644
From: zhongjuzhe
This patche add the testcases that are missed in v1.
*** BLURB HERE ***
zhongjuzhe (34):
RISC-V: Add vlex_1.c
RISC-V: Add vlex_2.c
RISC-V: Add vlex_1.C
RISC-V: Add mask load store testcases
RISC-V: Add vlexff_1.c
RISC-V: Add vlexff_2.c
RISC-V: Add vloxeix_1.c
From: zhongjuzhe
This patch is supplemental patch for [PATCH 14/21] which is missed in v1.
gcc/ChangeLog:
* config/riscv/constraints.md (vi): New constraint.
(vj): New constraint.
(vk): New constraint.
(vc): New constraint.
(Wn5): New constraint.
Hi!
On Mon, May 30, 2022 at 06:12:26PM +0800, Kewen.Lin wrote:
> on 2022/5/26 15:35, HAO CHEN GUI wrote:
> > This patch fixes the ICE reported in PR100736. It removes the condition
> > check of finite math only flag not setting in "*_cc" pattern.
> > With or without this flag, we still can use
On Wed, Jun 1, 2022 at 12:40 AM Richard Sandiford
wrote:
>
> Vladimir Makarov via Gcc-patches writes:
> > On 2022-05-29 23:05, Hongtao Liu wrote:
> >> On Fri, May 27, 2022 at 5:12 AM Vladimir Makarov via Gcc-patches
> >> wrote:
> >>>
> >>> On 2022-05-24 23:39, liuhongt wrote:
> Rigt now,
PING.
If there is anything else that needs to be addressed please let me know.
Sören Tempel wrote:
> I modified your patch to also define libgo_off_t_type (along to
> libgo_loff_t_type) and used that to define Offset_t in mksysinfo.sh.
> Furthermore, I fixed the include for the loff_t feature
A few globs missed in the .c -> .cc transition. Some targets were looking
at both *.c and *.cc, but there are no longer any .c files to scan.
Applying as obvious.
gcc/ChangeLog:
* Makefile.in (TAGS): Look at libcpp/*.cc.
gcc/c/ChangeLog:
* Make-lang.in (c.tags): Look at *.cc.
On 5/31/22 13:47, Patrick Palka wrote:
On Tue, 31 May 2022, Patrick Palka wrote:
r12-5487-g9bf69a8558638c replaced uses of timevar_cond_push/pop with
auto_cond_timevar and removed now unnecessary wrapper functions. This
patch does the same for timevar_push/pop and auto_timevar.
Bootstrapped
On 5/31/22 12:41, Patrick Palka wrote:
On Wed, 18 May 2022, Jason Merrill wrote:
On 5/17/22 12:34, Patrick Palka wrote:
On Sat, May 7, 2022 at 5:18 PM Jason Merrill wrote:
On 5/6/22 16:46, Patrick Palka wrote:
On Fri, 6 May 2022, Jason Merrill wrote:
On 5/6/22 16:10, Patrick Palka
On 5/31/22 13:13, Patrick Palka wrote:
This replaces the two differently named versions of the same function
with a single function using a default function argument.
Bootstrapped and regtested on x86_64-pc-linux-gnu, does this look OK
for trunk?
OK.
gcc/cp/ChangeLog:
* cp-tree.h
On 5/31/22 12:28, Patrick Palka wrote:
Here the out-of-line definition of Z::z causes duplicate_decls to
change z's type to use the implicit instantiation Z rather than the
corresponding primary template type (which is also the type of the
injected class name), and the former, being a dependent
On 5/31/22 08:56, Patrick Palka wrote:
On Sun, 29 May 2022, Jason Merrill wrote:
On 5/29/22 22:10, Jason Merrill wrote:
On 5/27/22 14:05, Patrick Palka wrote:
This makes us avoid substituting into the TEMPLATE_PARM_CONSTRAINT of
each template parameter except as necessary for (friend)
On Tue, 31 May 2022, Patrick Palka wrote:
> r12-5487-g9bf69a8558638c replaced uses of timevar_cond_push/pop with
> auto_cond_timevar and removed now unnecessary wrapper functions. This
> patch does the same for timevar_push/pop and auto_timevar.
>
> Bootstrapped and regtested on
r12-5487-g9bf69a8558638c replaced uses of timevar_cond_push/pop with
auto_cond_timevar and removed now unnecessary wrapper functions. This
patch does the same for timevar_push/pop and auto_timevar.
Bootstrapped and regtested on x86_64-pc-linux-gnu, does this look OK for
trunk?
gcc/cp/ChangeLog:
This replaces the two differently named versions of the same function
with a single function using a default function argument.
Bootstrapped and regtested on x86_64-pc-linux-gnu, does this look OK
for trunk?
gcc/cp/ChangeLog:
* cp-tree.h (cp_build_qualified_type_real): Rename to ...
On Mon, May 30, 2022 at 09:35:43AM +0100, Richard Sandiford wrote:
> "H.J. Lu" writes:
> > ---
> > RTL DSE tracks redundant constant stores within a basic block. When RTL
> > loop invariant motion hoists a constant initialization out of the loop
> > into a separate basic block, the constant
> Just checking there is still interest in this
Definitely, I am waiting for this to be able to send a new patch upstream
Cheers,
Tamar.
> -Original Message-
> From: Gcc-patches bounces+tamar.christina=arm@gcc.gnu.org> On Behalf Of Joel Hutton
> via Gcc-patches
> Sent: Tuesday,
On Wed, 18 May 2022, Jason Merrill wrote:
> On 5/17/22 12:34, Patrick Palka wrote:
> > On Sat, May 7, 2022 at 5:18 PM Jason Merrill wrote:
> > >
> > > On 5/6/22 16:46, Patrick Palka wrote:
> > > > On Fri, 6 May 2022, Jason Merrill wrote:
> > > >
> > > > > On 5/6/22 16:10, Patrick Palka wrote:
On 5/31/22 02:21, Aldy Hernandez wrote:
On Mon, May 30, 2022 at 4:56 PM Andrew MacLeod wrote:
On 5/30/22 09:27, Aldy Hernandez wrote:
Now that we have generic ranges, we need a way to define generic local
temporaries on the stack for intermediate calculations in the ranger
and elsewhere. We
Vladimir Makarov via Gcc-patches writes:
> On 2022-05-29 23:05, Hongtao Liu wrote:
>> On Fri, May 27, 2022 at 5:12 AM Vladimir Makarov via Gcc-patches
>> wrote:
>>>
>>> On 2022-05-24 23:39, liuhongt wrote:
Rigt now, mem_cost for separate mem alternative is 1 * frequency which
is pretty
Hi,
As of gdc-12, the lexer expects there 4 bytes of zero padding at the end
of the source buffer to mark the end of input. Sometimes when reading
from stdin, the data at the end of input is garbage rather than zeroes.
Fix that by explicitly calling memset past the end of the buffer.
On 2022-05-29 23:05, Hongtao Liu wrote:
On Fri, May 27, 2022 at 5:12 AM Vladimir Makarov via Gcc-patches
wrote:
On 2022-05-24 23:39, liuhongt wrote:
Rigt now, mem_cost for separate mem alternative is 1 * frequency which
is pretty small and caused the unnecessary SSE spill in the PR, I've
Here the out-of-line definition of Z::z causes duplicate_decls to
change z's type to use the implicit instantiation Z rather than the
corresponding primary template type (which is also the type of the
injected class name), and the former, being a dependent specialization,
lacks a TYPE_BINFO
My r13-680-g0dc8e1e7026d9b commit to add support for Decimal Floating
Point introduced:
case SDmode:
case DDmode:
case TDmode:
which are rejected by gcc-4.8 as build compiler.
This patch replaces them with E_SDmode, E_DDmode and E_TD_mode.
Committed as obvious.
*
This typo was fixed a little while ago in binutils-gdb with commit
e951225303. I noticed the difference today when importing libiberty
from gcc. Committed as obvious.
include/
* dwarf2.def: Correct spelling of DW_AT_namelist_item.
gcc/
* dwarf2out.cc (gen_namelist_decl): Adjust
On May 30, 2022, Richard Biener wrote:
> I don't think you can rely on TREE_VISITED not set at the start of the
> pass (and you don't clear it either).
I don't clear it, but I loop over all SSA names and set TREE_VISITED to
either true or false, so that's covered.
I even had a test patch that
On Sun, 29 May 2022, Jason Merrill wrote:
> On 5/29/22 22:10, Jason Merrill wrote:
> > On 5/27/22 14:05, Patrick Palka wrote:
> > > This makes us avoid substituting into the TEMPLATE_PARM_CONSTRAINT of
> > > each template parameter except as necessary for (friend) declaration
> > > matching, like
On Mon, 23 May 2022 at 22:57, Prathamesh Kulkarni
wrote:
>
> On Mon, 9 May 2022 at 21:21, Prathamesh Kulkarni
> wrote:
> >
> > On Mon, 9 May 2022 at 19:22, Richard Sandiford
> > wrote:
> > >
> > > Prathamesh Kulkarni writes:
> > > > On Tue, 3 May 2022 at 18:25, Richard Sandiford
> > > >
On Thu, 12 May 2022 at 16:15, Richard Sandiford
wrote:
>
> Prathamesh Kulkarni writes:
> > On Wed, 11 May 2022 at 12:44, Richard Sandiford
> > wrote:
> >>
> >> Prathamesh Kulkarni writes:
> >> > On Fri, 6 May 2022 at 16:00, Richard Sandiford
> >> > wrote:
> >> >>
> >> >> Prathamesh Kulkarni
From: zhongjuzhe
This patch adds implementation which missed in the V1 patch.
*** BLURB HERE ***
zhongjuzhe (1):
Add unit-stride load store intrinsics
.../riscv/riscv-vector-builtins-functions.cc | 80 +++
.../riscv/riscv-vector-builtins-functions.def | 7 ++
From: zhongjuzhe
gcc/ChangeLog:
* config/riscv/riscv-vector-builtins-functions.cc
(loadstore::assemble_name): New function.
(loadstore::get_argument_types): New function.
(vle::call_properties): New function.
(vle::get_return_type): New function.
> Can you post an updated patch (after the .cc renaming, and code_helper
> now already moved to tree.h).
>
> Thanks,
> Richard.
Patches attached. They already incorporated the .cc rename, now rebased to be
after the change to tree.h
Joel
0001-Refactor-to-allow-internal_fn-s.patch
On Mon, May 30, 2022 at 07:23:55PM +0200, Jakub Jelinek via Gcc-patches wrote:
> On Mon, May 30, 2022 at 10:43:30PM +0800, Chung-Lin Tang wrote:
> > > This feels like you only accept a single allocator in the new syntax,
> > > but that isn't my reading of the spec, I'd understand it as:
> > >
Hi!
OpenMP 5.2 adds support for firstprivate and allocate clauses on the scope
construct and this patch adds that support to GCC.
5.2 unfortunately (IMNSHO mistakenly) marked scope construct as worksharing,
which implies that it isn't possible to nest inside of it other scope,
worksharing loop,
> Am 31.05.2022 um 11:33 schrieb Jakub Jelinek via Gcc-patches
> :
>
> Hi!
>
> The following testcase ICEs because we use different types in comparison,
> idx has int type, while CASE_LOW has char type.
>
> While I believe all CASE_{LOW,HIGH} in the same switch have to use the same
> or
Hi!
The following testcase ICEs because we use different types in comparison,
idx has int type, while CASE_LOW has char type.
While I believe all CASE_{LOW,HIGH} in the same switch have to use the same
or compatible type, the index expression can have a promoted type as happens
in this testcase.
From: zhongjuzhe
gcc/ChangeLog:
* config/riscv/riscv-vector.cc (rvv_adjust_frame): Adjust frame
manipulation for RVV scalable vector.
* config/riscv/riscv-vector.h (rvv_adjust_frame): Adjust frame
manipulation for RVV scalable vector.
* config/riscv/riscv.cc
From: zhongjuzhe
gcc/ChangeLog:
* config/riscv/riscv-vector-builtins-functions.cc
(vset::assemble_name): New function.
(vset::get_argument_types): New function.
(vset::expand): New function.
(vget::assemble_name): New function.
From: zhongjuzhe
gcc/ChangeLog:
* config/riscv/riscv.cc (struct riscv_arg_info): Add calling convention
support.
(riscv_get_arg_info): Add calling convention support.
(riscv_function_arg_advance): Add calling convention support.
(riscv_pass_by_reference): Add
From: zhongjuzhe
gcc/ChangeLog:
* config/riscv/riscv-protos.h (rvv_expand_const_vector): New function.
(rvv_expand_const_mask): New function.
(rvv_const_vec_all_same_in_range_p): New function.
* config/riscv/riscv-vector.cc (classify_vtype_field): Move codes
From: zhongjuzhe
gcc/ChangeLog:
* config/riscv/riscv-protos.h (riscv_add_offset): Change
riscv_add_offset as global function.
* config/riscv/riscv-vector.cc (rvv_report_required): New function.
(expand_quotient): New function.
(rvv_expand_poly_move): New
From: zhongjuzhe
gcc/ChangeLog:
* config.gcc: Add riscv-insert-vsetvl.o extra_objs for RVV support.
* config/riscv/constraints.md (Ws5): New constraint.
* config/riscv/predicates.md (p_reg_or_const_csr_operand): New
predicate.
(vector_reg_or_const0_operand): New
From: zhongjuzhe
gcc/ChangeLog:
* config/riscv/riscv-protos.h (riscv_register_pragmas): New function.
(riscv_classify_vlmul_field): New enum.
(rvv_classify_vlmul_field): New enum.
(rvv_parse_vsew_field): New enum.
(rvv_parse_vlmul_field): New enum.
From: zhongjuzhe
gcc/ChangeLog:
* config.gcc: New header.
* config/riscv/riscv-c.cc (riscv_pragma_intrinsic): New function.
(riscv_check_builtin_call): New function.
(riscv_register_pragmas): New function.
* config/riscv/riscv-protos.h
From: zhongjuzhe
gcc/ChangeLog:
* config/riscv/riscv-vector-builtins.cc (make_type_sizeless): New
function.
(sizeless_type_p): New function.
(vector_builtin_mode): New function.
(vector_legal_vlmul): New function.
(add_vector_type_attribute): New
From: zhongjuzhe
gcc/ChangeLog:
* config.gcc: Add riscv-vector-builtins-functions.o and
riscv-vector-builtins.o extra_objs for RVV support.
* config/riscv/riscv-builtins.cc (riscv_init_builtins): Add RVV support.
(riscv_builtin_decl): Add RVV support.
From: zhongjuzhe
gcc/ChangeLog:
* config.gcc: Add riscv-vector.o extra_objs for RVV support.
* config/riscv/constraints.md: New constraints.
* config/riscv/predicates.md: New predicates.
* config/riscv/riscv-modes.def: New machine mode.
*
On Mon, May 30, 2022 at 4:56 PM Andrew MacLeod wrote:
>
> On 5/30/22 09:27, Aldy Hernandez wrote:
> > Now that we have generic ranges, we need a way to define generic local
> > temporaries on the stack for intermediate calculations in the ranger
> > and elsewhere. We need temporaries analogous
57 matches
Mail list logo
