m not sure how to feel about merging complexity into the cost part
of an address cost, though.
> If it's really only about the "complexity" value then each
> compensation step should
> add to the complexity?
That could be the way to go. Also worth verifying is that we compensate f
ds,
Dimitrije
From: Richard Biener
Sent: Monday, November 7, 2022 2:35 PM
To: Dimitrije Milosevic
Cc: Jeff Law ; gcc-patches@gcc.gnu.org
; Djordje Todorovic
Subject: Re: [PATCH 1/2] ivopts: Revert computation of address cost complexity.
On Wed, Nov 2, 2022 at 9:40 AM Dimitrije Milosevic
wro
'll take a deeper look into the candidate selection algorithm then. Will
get back to you.
Regards,
Dimitrije
From: Jeff Law
Sent: Tuesday, November 1, 2022 7:46 PM
To: Richard Biener; Dimitrije Milosevic
Cc: gcc-patches@gcc.gnu.org; Djordje Todorovic
l
Regards,
Dimitrije
From: Richard Biener
Sent: Friday, October 28, 2022 9:38 AM
To: Dimitrije Milosevic
Cc: gcc-patches@gcc.gnu.org ; Djordje Todorovic
Subject: Re: [PATCH 2/2] ivopts: Consider number of invariants when calculating
register pressure.
On Tue, Oct 25, 2022 at 3:00 PM
n.
Would love to hear your thoughts.
[0] https://gcc.gnu.org/pipermail/gcc-patches/2022-October/604304.html
Regards,
Dimitrije
From: Richard Biener
Sent: Friday, October 28, 2022 9:00 AM
To: Dimitrije Milosevic
Cc: Jeff Law ; gcc-patches@gcc.gnu.org
; Djordje Todorovic
Subject: Re: [PATCH
s to compensate
the lack of other addressing modes somehow. If complexities for both of
those are equal to 0, in cases where complexities decide which candidate is
to be chosen, a more complex candidate may be picked.
Regards,
Dimitrije
From: Jeff Law
Sent: Friday, October 28, 2022 1:02 AM
To: D
c1$f20,-8($2)
lwc1$f12,0($25)
maddf.s $f11,$f1,$f12
bne $2,$20,.L93
swc1$f11,-4($2)
...
= After f9f69dd =
Notice the additional instructions used for index calculation, due to
unoptimal candidate selection.
Regards,
Dimitrije
rd Biener
Sent: Tuesday, October 25, 2022 1:07 PM
To: Dimitrije Milosevic
Cc: gcc-patches@gcc.gnu.org ; Djordje Todorovic
Subject: Re: [PATCH 2/2] ivopts: Consider number of invariants when calculating
register pressure.
On Fri, Oct 21, 2022 at 3:57 PM Dimitrije Milosevic
wrote:
>
>
(ivopts_estimate_reg_pressure): Adjust.
Signed-off-by: Dimitrije Milosevic
---
gcc/tree-ssa-loop-ivopts.cc | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/gcc/tree-ssa-loop-ivopts.cc b/gcc/tree-ssa-loop-ivopts.cc
index d53ba05a4f6..9d0b669d671 100644
--- a/gcc/tree-ssa
-ivopts.cc (compute_symbol_and_var_present): Reintroduce.
(compute_min_and_max_offset): Likewise.
(get_address_cost): Revert
complexity calculation.
Signed-off-by: Dimitrije Milosevic
---
gcc/tree-ssa-address.cc | 2 +-
gcc/tree-ssa-address.h | 2 +
gcc/tree-ssa-loop
nested
loops.
Additionally, when bumping up the register pressure cost, the number of
invariants should also be
considered, in addition to the number of candidates.
Dimitrije Milosevic (2):
ivopts: Revert computation of address cost complexity.
ivopts: Consider number of invariants when calcu
Gentle ping. :)
From: Dimitrije Milosevic
Sent: Friday, July 29, 2022 12:38 PM
To: gcc-patches@gcc.gnu.org
Cc: Djordje Todorovic ; xry...@xry111.site
; mask...@google.com
Subject: [PATCH] Mips: Enable TSAN for 64-bit ABIs
The following patch enables TSAN for mips64, on which it is supported
The following patch enables TSAN for mips64, on which it is supported.
Signed-off-by: Dimitrije Milosevic .
libsanitizer/ChangeLog:
* configure.tgt: Enable
TSAN for 64-bit ABIs.
---
libsanitizer/configure.tgt | 4
1 file changed, 4 insertions(+)
diff --git a/libsanitizer
Thanks Martin! I'm sending out the output from git format-patch as an
attachment to this email.
From: Martin Liška
Sent: Thursday, July 28, 2022 3:48 PM
To: Dimitrije Milosevic ;
gcc-patches@gcc.gnu.org
Cc: ma...@embecosm.com ; Djordje Todorovic
; jos...@codesourcery.com
Subjec
Gentle ping, requiring someone to push the change. :)
From: Dimitrije Milosevic
Sent: Monday, July 25, 2022 8:55 AM
To: gcc-patches@gcc.gnu.org
Cc: Djordje Todorovic ; xry...@xry111.site
Subject: [PATCH] libsanitizer: Cherry-pick 2bfb0fcb51510f22723c8cdfefe from
upstream
> Do you know someone very familiar with MIPS and GCC and capable as a
> port maintainer? An active MIPS port maintainer will make the situation
> better.
Sadly, no. I agree it would make things easier.
From: Xi Ruoyao
Sent: Wednesday, July 27, 2022 8:43 AM
To: Dimitrije Milosev
Gentle ping, requiring someone to push this change, as I do not have commit
access. :)
From: Dimitrije Milosevic
Sent: Monday, July 25, 2022 8:55 AM
To: gcc-patches@gcc.gnu.org
Cc: Djordje Todorovic ; xry...@xry111.site
Subject: [PATCH] libsanitizer: Cherry-pick 2bfb0fcb51510f22723c8cdfefe
2bfb0fcb51510f22723c8cdfefe [Sanitizer][MIPS] Fix stat struct size for the O32
ABI.
Signed-off-by: Dimitrije Milosevic .
---
.../sanitizer_common/sanitizer_platform_limits_posix.h | 7 ---
1 file changed, 4 insertions(+), 3 deletions(-)
diff --git a/libsanitizer/sanitizer_common
an Glibc (it's PR
> 106136 in GCC bugzilla).
I might be able to, but no promises. :)
From: Xi Ruoyao
Sent: Wednesday, July 13, 2022 4:38 AM
To: Dimitrije Milosevic ; Hans-Peter Nilsson
Cc: Djordje Todorovic ; gcc-patches@gcc.gnu.org
Subject: Re: Mips: Fix kernel_stat structure size
s
will be working properly. As for this change, I'm not sure what should be done?
Should this be committed now, while the LLVM change is cherry-picked once it's
committed.
Best regards,
Dimitrije Milosevic
From: Hans-Peter Nilsson
Sent: Saturday, July 9, 2022 4:44 PM
To: Xi Ruoyao
Cc:
1
-# if defined(__mips64)
+# if defined(__mips64) && _MIPS_SIM == _ABI64
#define SANITIZER_MIPS32 0
#define SANITIZER_MIPS64 1
# else
--
2.25.1
From: Richard Sandiford
Sent: Monday, July 4, 2022 1:23 PM
To: Xi Ruoyao via Gcc-patches
Cc: Dimitrije Milosevic ; Xi Ruoyao
; Djordje
Ping. :)
I think this is good to go. Unfortunately, I do not have commit access, so if
anyone
can commit it, that would be great!
From: Dimitrije Milosevic
Sent: Friday, July 1, 2022 4:25 PM
To: Xi Ruoyao ; gcc-patches@gcc.gnu.org
Cc: Djordje Todorovic ; Richard Sandiford
Subject: Re: Mips
flag_asynchronous_unwind_tables = 1;
From: Xi Ruoyao
Sent: Tuesday, July 5, 2022 3:54 AM
To: Dimitrije Milosevic ;
gcc-patches@gcc.gnu.org
Cc: Djordje Todorovic
Subject: Re: [PATCH] Mips: Enable asynchronous unwind tables with both ASAN and
TSAN
On Mon, 2022-07-04 at 14:28 +, Di
On Saturday, June 11, 2022 2:03 PM, Xi wrote:
> Just tried TSAN_SUPPORTED=yes with asynchronous unwind tables enabled,
> but I got some strange test failures for tls_race.c:
>
> FAIL: c-c++-common/tsan/tls_race.c -O0 output pattern test
> Output was:
> ThreadSanitizer: CHECK failed: tsan_platfo
Thanks Xi. Forgive me as I'm not that familiar with the coding standards
when submitting patches for a review.
Here is the updated version of the patch.
Building the ASAN for the n32 MIPS ABI currently fails, due to a few reasons:
- defined(__mips64), which is set solely based on the architecture
Thanks Xi. Forgive me as I'm not that familiar with the coding standards
when submitting patches for a review.
Here is the updated version of the patch.
Fix kernel_stat structure size for non-Android 32-bit Mips.
LLVM currently has this value for the kernel_stat structure size,
as per compiler-rt
When we use cross toolchains, in which the GCC libraries are not installed
within a designated system root, the shared sanitizer libraries link against
libstdc++.so* within the same libraries. This directory, however, is not in
RPATH,
so attempting to build a dynamically linked application with
Building the ASAN for the n32 MIPS ABI currently fails, due to a few reasons:
- defined(__mips64), which is set solely based on the architecture type
(32-bit/64-bit),
was still used in some places. Therefore, defined(__mips64) is swapped with
SANITIZER_MIPS64,
which takes the ABI into account a
Fix kernel_stat structure size for non-Android 32-bit Mips.
LLVM currently has this value for the kernel_stat structure size,
as per compiler-rt/lib/sanitizer-common/sanitizer_platform_limits_posix.h.
This also resolves one of the build issues for non-Android 32-bit Mips.
libsanitizer/ChangeLo
When we use cross toolchains, in which the GCC libraries are not installed
within a designated system root, the shared sanitizer libraries link against
libstdc++.so* within the same libraries. This directory, however, is not in
RPATH, so attempting to build a dynamically linked application with
Building the ASAN for the n32 MIPS ABI currently fails, due to a few reasons:
- defined(__mips64), which is set solely based on the architecture type
(32-bit/64-bit), was still used in some places.
Therefore, defined(__mips64) is swapped with SANITIZER_MIPS64, which takes the
ABI into account as
Definitely, a patch is on the way.
From: Xi Ruoyao
Sent: Tuesday, June 7, 2022 10:20 AM
To: Dimitrije Milosevic ;
gcc-patches@gcc.gnu.org
Cc: Djordje Todorovic
Subject: Re: [PATCH] Mips: Enable asynchronous unwind tables with both ASAN and
TSAN
On Mon, 2022
Correct, it should be committed very soon.
From: Xi Ruoyao
Sent: Tuesday, June 7, 2022 10:17 AM
To: Dimitrije Milosevic ;
gcc-patches@gcc.gnu.org
Cc: Djordje Todorovic
Subject: Re: [PATCH] Mips: Fix the ASAN shadow offset hook for the n32 ABI
On Mon, 2022-06
Fix the ASAN shadow offset hook for the n32 ABI.
gcc/ChangeLog:
* config/mips/mips.cc (mips_asan_shadow_offset): Reformat
to handle the N32 ABI.
* config/mips/mips.h (SUBTARGET_SHADOW_OFFSET): Remove
the macro, as it is not needed anymore.
---
gcc/config/mips/mi
&& (flag_sanitize & SANITIZE_THREAD)))
&& !global_options_set.x_flag_asynchronous_unwind_tables)
flag_asynchronous_unwind_tables = 1;
---
From: Xi Ruoyao
Sent: Saturday, May 28, 2022 12:30 PM
To: Dimitrije Milosevic ;
gcc-patches@gcc.gnu.o
Enable asynchronous unwind tables with both ASAN and TSAN for correct
back-trace.
LLVM currently enables asynchronous unwind tables for: ASAN, HWSAN, TSAN, MSAN,
and DFSAN.
HWSAN is currently available only on AArch64, while MSAN and DFSAN are not
available at all.
Also, LLVM checks is '-ffreest
36 matches
Mail list logo
