Steve Ellcey wrote:
> On Mon, 2013-08-19 at 22:21 -0700, Teresa Johnson wrote:
>
>> 2013-08-19 Teresa Johnson
>>
>> PR rtl-optimizations/57451
>> * final.c (reemit_insn_block_notes): Prevent lexical blocks
>> from crossing split section boundaries.
>>
>> * tes
Christian Bruel wrote:
> The attached patch fixes an ice while building the linux kernel. Reduced
> in the included testcase.
>
> The problem is that we are generating a movhi_reg_reg insn that accepts
> only registers as operands. Spilling a pseudo on the stack results in an
> invalid memory loa
Christian Bruel wrote:
> && (!can_create_pseudo_p () && REG_P (operands[0]) && REG_P (operands[1]))"
>
> is necessary ?
It looks an another hack to allow the 2nd and 3rd alternatives only
when reloading. If so, it might be a bit cleaner to use a special
predicate like
;; Returns 1 if OP can be
Christian Bruel wrote:
> This patch fixes the aforementioned PR by refusing FPUL_REG to be an
> acceptable reg for any arithmetic_operand on TARGET_SH4. (This was a
> strange SH4 singularity with regards to the SH family).
>
> The only impacted insn is movsf_ie used for reg-fpreg transfers. So th
Christian Bruel wrote:
> (define_insn "*mov_reg_reg"
> - [(set (match_operand:QIHI 0 "arith_reg_dest" "=r,m,*z")
> - (match_operand:QIHI 1 "register_operand" "r,*z,m"))]
> - "TARGET_SH1 && !t_reg_operand (operands[1], VOIDmode)"
> + [(set (match_operand:QIHI 0 "general_movdst_operand" "=r,
> Christian Bruel wrote:
>> (define_insn "*mov_reg_reg"
>> - [(set (match_operand:QIHI 0 "arith_reg_dest" "=r,m,*z")
>> -(match_operand:QIHI 1 "register_operand" "r,*z,m"))]
>> - "TARGET_SH1 && !t_reg_operand (operands[1], VOIDmode)"
>> + [(set (match_operand:QIHI 0 "general_movdst_operand
Oleg Endo wrote:
> This implements the changes as proposed PR, albeit with some small
> differences:
> --- gcc/config/sh/sh.c(revision 191865)
> +++ gcc/config/sh/sh.c(working copy)
[snip]
> + std::vector tokens;
> + for (std::stringstream ss (str); ss.good (); )
> + {
> +t
Oleg Endo wrote:
> The existing .c files are compiled as C++ already. There was a
> discussion not long go whether the .c files should be renamed to .cc or
> not. If I remember correctly, the conclusion was that existing .c files
> remain .c, while files newly added should be .cc. gcc/double-in
Gabriel Dos Reis wrote:
>> Ah, I've missed that argument. Then can we use STL classes in .c now?
>
> Yes.
>
> The existing file extensions ".c" do not mean "C" only.
Thanks for clarification.
Oleg, the patch is OK.
Regards,
kaz
Oleg Endo wrote:
> This handles the case where the T bit is stored to a reg as the value
> 0x7FFF or 0x8000.
> Tested on rev 191894 with
> make -k check RUNTESTFLAGS="--target_board=sh-sim
> \{-m2/-ml,-m2/-mb,-m2a/-mb,-m4/-ml,-m4/-mb,-m4a/-ml,-m4a/-mb}"
>
> and no new failures.
> OK?
OK.
Oleg Endo wrote:
> This is the patch as proposed in the PR to make
> libgcc/config/sh/linux-atomic use the appropriate compiler generated
> atomic built-in functions depending on the currently selected
> atomic-model.
>
> Tested on 191894 with 'make all-gcc' and by compiling code to see if the
>
Oleg Endo wrote:
> This adds the two common built-in functions __builtin_thread_pointer and
> __builtin_set_thread_pointer to the SH port.
> I've done it in a way so that hopefully it can be transitioned to target
> independent thread pointer built-ins easily, as suggested by Richard a
> while ago
Hi,
I've committed the patch below to fix sh64-elf build failure.
SHmedia and SHcompact using call cookie require special return insns
and will require an extra work to enable simple_return. I simply
disable it for these targets ATM.
Regards,
kaz
--
2012-10-04 Kaz K
Oleg Endo wrote:
> Since the -mieee behavior has been fixed, is enabled by default on SH
> and the additional flags in libgcc can be removed.
>
> OK?
OK.
Regards,
kaz
Oleg Endo wrote:
> The atomic options of SH have been changed recently. The attached patch
> updates the 4.8 changes.html accordingly, plus some minor wording fixes.
>
> OK?
OK.
Regards,
kaz
Oleg Endo wrote:
> Do you mean something like the attached patch as a preparation step?
> (checked with 'make all')
Yes. The patch is OK with removing the first line of the ChangeLog
entry for PR number.
Regards,
kaz
Oleg Endo wrote:
> The attached patch is the next step that adds the thread pointer
> builtins. The GBR address mode stuff will follow afterwards separately.
> Tested on rev 192142 with 'make all' and
> 'make -k check-gcc RUNTESTFLAGS="sh.exp --target_board=sh-sim
> \{-m2/-ml,-m2/-mb,-m2a/-mb,-m4
Oleg Endo wrote:
> Attached patch adds the GBR addressing mode stuff. I've added the
> '!NONJUMP_INSN_P (i)' check which should fix the crash. Could you
> please test again?
No new failures with the top level "make -k check" on cross
sh4-unknown-linux-gnu. OK for trunk.
Regards,
kaz
Oleg Endo wrote:
> The attached patch improves comparisons such as
> 'unsigned int <= 0x7FFF' on SH.
> As mentioned in the PR, for some reason, those comparisons do not go
> through the cstore expander. As a consequence the comparison doesn't
> get the chance to be canonicalized by the target
Oleg Endo wrote:
> I've just checked this against current trunk (rev 192193). The problem
> seems to be gone. There's also a total decrease of 152 bytes on this
> file without the patch. So it seems it was a different issue. The diff
> is now:
Thanks for looking into it quickly. The patch is
Oleg Endo wrote:
> This adds documentation on the new thread pointer built-ins that were
> added recently to the SH target.
> Tested with 'make info dvi pdf'.
> OK?
OK.
Regards,
kaz
Oleg Endo wrote:
> This adds the reduced test case as mentioned by Kaz in the PR to the
> test suite.
> Tested with
> make -k check-gcc RUNTESTFLAGS="compile.exp=pr34777*
> --target_board=sh-sim
> \{-m2/-ml,-m2/-mb,-m2a/-mb,-m4/-ml,-m4/-mb,-m4a/-ml,-m4a/-mb}"
>
> OK?
It should be put into gcc.ta
Oleg Endo wrote:
> This documents the new thread pointer built-ins in the SH www changes
> for 4.8.
> OK?
Looks OK to me.
Regards,
kaz
Oleg Endo wrote:
> Uhm, yes, I forgot to add the -fschedule-insns and -mprefergot options.
> Regarding the -Os option, I think it's better to test this one at
> multiple optimization levels, just in case. I've looked through
> gcc.c-torture/compile and found some target specific test cases there,
Oleg Endo wrote:
> This is the same patch I posted in the PR. It seems to fix the issue.
> Tested on rev 192200 with
> make -k check RUNTESTFLAGS="--target_board=sh-sim
> \{-m2/-ml,-m2/-mb,-m2a/-mb,-m4/-ml,-m4/-mb,-m4a/-ml,-m4a/-mb}"
>
> and no new failures.
> OK?
OK.
Regards,
kaz
Oleg Endo wrote:
> This one further improves T bit stores and conditional branches on SH
> for cases like described in comment #53 in the PR.
> Tested on rev 192200 with
> make -k check RUNTESTFLAGS="--target_board=sh-sim
> \{-m2/-ml,-m2/-mb,-m2a/-mb,-m4/-ml,-m4/-mb,-m4a/-ml,-m4a/-mb}"
>
> and no
Oleg Endo wrote:
> Anyway, regardless of this failure, the attached patch for SH should be
> applicable. OK?
OK.
Regards,
kaz
Oleg Endo wrote:
> This fixes the issue of PR 54602 as proposed in the PR.
> Tested on rev 192200 with
> make -k check RUNTESTFLAGS="--target_board=sh-sim
> \{-m2/-ml,-m2/-mb,-m2a/-mb,-m4/-ml,-m4/-mb,-m4a/-ml,-m4a/-mb}"
>
> and no new failures.
> OK?
OK.
Regards,
kaz
Oleg Endo wrote:
> The attached patch fixes PR 54680.
> Tested on rev 192200 with
> make -k check RUNTESTFLAGS="--target_board=sh-sim
> \{-m2/-ml,-m2/-mb,-m2a/-mb,-m4/-ml,-m4/-mb,-m4a/-ml,-m4a/-mb}"
>
> and no new failures.
> OK?
OK.
Regards,
kaz
Oleg Endo wrote:
> The attached patch adds documentation for SH specific function
> attributes which haven't been documented yet.
> Tested with 'make info dvi pdf'.
> OK?
OK.
Regards,
kaz
Oleg Endo wrote:
> The attached patch adds gcc.target/sh/torture and puts the test there.
> The torture subdir might be also useful in the future.
> Tested on rev 192417 with
> make -k check-gcc RUNTESTFLAGS="--target_board=sh-sim\{-m2/-ml}"
>
> OK?
OK.
Regards,
kaz
Oleg Endo wrote:
> I somehow initially forgot to implement DImode GBR based loads/stores.
> Attached patch does that and also fixes a problem with the GBR address
> mode optimization.
> Tested on rev 192417 with
> make -k check RUNTESTFLAGS="--target_board=sh-sim
> \{-m2/-ml,-m2/-mb,-m2a/-mb,-m4/-
Oleg Endo wrote:
> This one refactors some copy pasta that my previous patch regarding this
> matter introduced and catches more unnecessary sign/zero extensions of T
> bit stores. It also fixes the bug reported in PR 54925 which popped up
> after the last patch for PR 51244.
> Tested on rev 1924
Oleg Endo wrote:
> This adds the test case from the PR.
> Tested together with the patch posted here
> http://gcc.gnu.org/ml/gcc-patches/2012-10/msg01380.html
>
> OK?
It would be better to make it a valid C program. I've checked
that the test case with the change below also ICEs on revision
192
Ralf Corsepius wrote:
> I would like to apply the patch below to trunk and gcc-4.7 branch.
> GCC doesn't build for sh-rtems* without it.
OK. Please go ahead.
Regards,
kaz
Oleg Endo wrote:
> This fixes the issues of PR 54988.
> Tested on rev 192482 with
> make -k check RUNTESTFLAGS="--target_board=sh-sim
> \{-m2/-ml,-m2/-mb,-m2a/-mb,-m4/-ml,-m4/-mb,-m4a/-ml,-m4a/-mb}"
>
> and no new failures.
OK.
Regards,
kaz
Oleg Endo wrote:
> Some of the things I've done in 4.8 to improve SH T bit handling turned
> out to produce wrong code. The attached patch fixes that by introducing
> an SH specific RTL pass.
>
> Tested on rev 202876 with
> make -k check RUNTESTFLAGS="--target_board=sh-sim
> \{-m2/-ml,-m2/-mb,-m
Oleg Endo wrote:
> Forgot to handle a case in function can_remove_cstore, thanks for
> catching it. Fixed in the attached patch and also added test cases.
> Retested as before without new failures.
Ok for trunk.
> Yeah, right. I've changed 'ifcvt_sh' to 'sh_ifcvt'.
>+ register_pass (make_pas
Oleg Endo wrote:
> To be honest, I had some difficulty picking the name.
> Maybe something like 'sh_tbit_combine' or 'sh_treg_combine' would be
> better, or at least less confusing? Suggestions are highly appreciated.
'sh_treg_combine' or 'sh_combine_treg' sounds good to me.
Regards,
ka
Christian Bruel wrote:
> This patch just reorganizes the SH code used for memory builtins into
> its own file, in preparation of the RTL strcmp hoisting in the next part.
>
> OK for trunk ?
A minor nit:
> * gcc/config/sh/sh-mem (expand_block_move): Moved here.
lost .c?
Ok with that chan
Christian Bruel wrote:
> thanks for having retested this, The tests are still not complete for
> RTL generated functions, there are cases where no str/cmp wil be
> emitted, because we can predict than the size is less than 4 and so have
> a direct fallthru into the byte at a byte loop copying.
>
Oleg Endo wrote:
> The attached patch fixes volatile mem stores on SH so that they won't
> result in redundant sign/zero extensions and will utilize available
> addressing modes. This is similar to what has been done to fix memory
> loads in http://gcc.gnu.org/ml/gcc-patches/2013-06/msg01315.html
Christian Bruel wrote:
> No regressions for -m2 and -m4 for sh-elf.
> OK for trunk ?
OK with the change of ChangeLog entry suggested by your another mail.
Thanks!
Regards,
kaz
Oleg Endo wrote:
> The attached patch adds a new macro TARGET_BIG_ENDIAN. I think it's a
> bit easier to read than !TARGET_LITTLE_ENDIAN. Moreover, some not so
> obvious looking uses of TARGET_LITTLE_ENDIAN for selecting MSW and LSW
> register offsets are clarified by using new macros SH_REG_MSW
Oleg Endo wrote:
> This adds some more patterns to utilize the SH addc instruction.
> Tested on rev 204111 with
> make -k check RUNTESTFLAGS="--target_board=sh-sim
> \{-m2a/-mb,-m2a-single/-mb,-m4/-ml,-m4-single/-ml,-m4/-mb,-m4-single/-mb}"
>
> and no new failures.
> OK for trunk?
Using same "*a
Oleg Endo wrote:
> This fixes a bug on the 4.8 branch which was initially described in PR
> 51244 comment 59 and also reported as a separate PR 59343.
> Tested with
> make -k check RUNTESTFLAGS="--target_board=sh-sim
> \{-m2/-ml,-m2/-mb,-m2a/-mb,-m4/-ml,-m4/-mb,-m4a/-ml,-m4a/-mb}"
>
> and no new
Richard Biener wrote:
> Note this opens the window for other important wrong-code fixes - please
> CC me on any you'd like to propose for GCC 5.2 and wait for my
> explicit approval.
Can I backport the patch in
[SH] Fix PR target/66780
https://gcc.gnu.org/ml/gcc-patches/2015-07/msg00472.html
to
Richard Biener wrote:
>> Can I backport the patch in
>>
>> [SH] Fix PR target/66780
>> https://gcc.gnu.org/ml/gcc-patches/2015-07/msg00472.html
>>
>> to gcc-5.2? It's a few lines change for a SH specific wrong code bug
>> with -fstack-protector which is disastrous for Debian SH folks who use
>>
I'll backport it to 4.9 later and
to 5 when it reopens.
Regards,
kaz
--
2015-07-16 Kaz Kojima
PR target/65249
* config/sh/sh.md (movdi): Split simple reg move to two movsi
when the destination is R0.
diff --git a/config/sh/sh.md b/config/sh/sh.md
index f0cb3c
I've committed the one liner below to fix PR target/67049 which causes
build failures for sh64-elf on 5/6. There is a typo in GOTaddr2picreg,
which is my bad. Tested on sh64-unknown-elf.
Regards,
kaz
--
2015-07-31 Kaz Kojima
PR target/67049
* config/sh/
Richard Sandiford wrote:
> This patch converts the code to do everything via the optabs interface.
> This required a change to the sh.md cmpstrn pattern, where the length
> predicate was "immediate_operand" but where the expander code wanted
> to handle nonimmediates too. I also removed the now-r
I've committed the attached patch to fix PR67573. LRA reveals
that scratch registers in *call_*pcrel should be defined as
early clobbered. The patch is tested on sh4-unkown-linux-gnu.
Regards,
kaz
--
2015-09-16 Kaz Kojima
PR target/67573
* config/sh/sh.md (call_
I've committed the attached obvious fix for build failure for SH.
object_allocator is changed so to remove the 2nd argument of its
constructor.
Regards,
kaz
--
2015-09-17 Kaz Kojima
* config/sh/sh.c (label_ref_list_d_pool): Adjust to
object_allocator change.
oblem. Tested on sh4-unknown-linux-gnu.
Committed as revision 227953.
Regards,
kaz
--
2015-09-21 Kaz Kojima
Backport from mainline
2015-09-16 Kaz Kojima
PR target/67573
* config/sh/sh.md (call_pcrel): Add early clobber to scratch
I've committed the attached patch to fix PR target/67716. It
implements targetm.override_options_after_change for SH. Tested
on sh4-unknown-linux-gnu.
Regards,
kaz
--
2015-09-29 Kaz Kojima
PR target/67716
* config/sh/sh.c (sh_override_options_after_change)
Rich Felker wrote:
> I worked around it and opened an issue for it:
>
> https://gcc.gnu.org/bugzilla/show_bug.cgi?id=67812
>
> But trying the patch on vanilla GCC trunk without my usual J2 target
> setup revealed some additional issues I need to address. I'm getting
> ICE in the code that genera
Oleg Endo wrote:
>> So apparently the strange behavior I observed is intended. Presumably
>> there is some mechanism to ensure that these functions are always
>> static-linked? But I don't see it. The libgcc spec I see is:
>>
>> *libgcc:
>> %{static|static-libgcc:-lgcc
>> -lgcc_eh}%{!static:%{!st
H targets. I'd like to
apply it anyway, because it fixes the ICEs and there are no
new failures on sh4-unknown-linux-gnu. Committed.
Regards,
kaz
--
2015-10-05 Kaz Kojima
PR c/65345
* config/sh/sh.c (sh_atomic_assign_expand_fenv): Adjust to use
create_tmp_var_r
Eric Botcazou wrote:
> No, it is not if you don't need to make the variable addressable. In any
> case, the failure mode is another ICE in make_decl_rtl so easy to spot.
Thanks for your explanation! It clarifies the intent of
the original i386 patch.
Regards,
kaz
Oleg Endo wrote:
> On Mon, 2015-10-05 at 23:15 -0400, Rich Felker wrote:
>> - Is the use of self specs (see DRIVER_SELF_SPECS in sh.h) an
>> acceptable way to set the default? I brought this up before but
>> don't think anyone answered. I find this method more clear and less
>> invasive (doe
kaz
--
2015-10-26 Kaz Kojima
PR target/68091
* config/sh/sh.c (sh_vector_mode_supported_p): Use
TARGET_SHMEDIA_FPU instead of TARGET_FPU_ANY.
diff --git a/gcc/config/sh/sh.c b/gcc/config/sh/sh.c
index f8187e4..a153845 100644
--- a/gcc/config/sh/sh.c
+++ b/gcc/config/sh
I've committed the attached obvious patch to fix
FAIL: compiler driver --help=target option(s): "^ +-.*[^:.]$" absent from
output: " -mfdpic Generate ELF FDPIC code"
popping up after SH fdpic patch applied.
Regards,
kaz
Rich Felker wrote:
> The "chk_guard_add" pattern used for loading the GOT slot address for
> __stack_chk_guard hard-codes use of r12 as a fixed GOT register and
> thus is not suitable for FDPIC, where the saved initial value of r12
> from function entry is what we need.
The patch is OK. Committe
15-12-06 Kaz Kojima
* config/sh/sh.md (rsqrtsf2): Adjust for canonical form with unspec.
diff --git a/config/sh/sh.md b/config/sh/sh.md
index 083febe..b6fe05c 100644
--- a/config/sh/sh.md
+++ b/config/sh/sh.md
@@ -13744,12 +13744,11 @@ label:
(define_insn "rsqrtsf2
Jan-Benedict Glaw wrote:
> shle-linux breaks with:
>
> g++ -fno-PIE -c -g -O2 -DIN_GCC -DCROSS_DIRECTORY_STRUCTURE
> -fno-exceptions -fno-rtti -fasynchronous-unwind-tables -W -Wall
> -Wno-narrowing -Wwrite-strings -Wcast-qual -Wmissing-format-attribute
> -Woverloaded-virtual -pedantic -Wn
62111
for details. Tested on sh64-elf and sh4-unknown-linux-gnu with
no new failures. I'll backport this to 4.9 in a week or so.
Regards,
kaz
--
2014-08-25 Kaz Kojima
PR target/62111
* config/sh/predicates.md (general_extend_operand): Disable
TRUNCATE bef
.
Regards,
kaz
--
2014-08-27 Kaz Kojima
PR target/62261
* config/sh/sh.md (ashlsi3): Handle negative shift count for
TARGET_SHMEDIA.
(ashldi3, ashrsi3, ashrdi3, lshrsi3, lshrdi3): Likewise.
--- ORIG/trunk/gcc/config/sh/sh.md 2014-07-03 09:10:37.0
Hi,
gcc.c-torture/execute/pr39228.c fails with "(test for excess errors)"
on SH for recent revisions. My gcc.log says:
gcc.c-torture/execute/pr39228.c:20:43: warning: always_inline function might
not be inlinable [-Wattributes]
...
It looks that alpha has the similar issue:
https://gcc.gnu.org
Oleg Endo wrote:
> -mieee should be the default on sh* and thus can be removed from the
> dg-options line, or is it not? If -mieee is still needed (for alpha) maybe
> it's better to use dg-additional-options instead?
Sure. The attached is a revised one.
Regards,
kaz
--
* gcc.
Uros Bizjak wrote:
> -/* { dg-options "-mieee" { target sh*-*-* alpha*-*-* } } */
> +/* { dg-options "-w -mieee" { target sh*-*-* alpha*-*-* } } */
> /* { dg-skip-if "No Inf/NaN support" { spu-*-* } "*" "" } */
>
> Please use /* { dg-add-options ieee } */ directive here. There is
> another one p
Jeff Law wrote:
>> * gcc.c-torture/execute/pr39228.c: Use dg-additional-options
>> instead of dg-options and remove sh*-*-* from its target list.
>> Add "inline" keyword to test functions.
> Wouldn't we be better off moving this into execute/ieee?
I've tried it and found that all e
28.c makes me think that this test
can be moved into execute/ieee. Thought?
Regards,
kaz
--
2014-09-04 Kaz Kojima
* gcc.c-torture/execute/pr44683.c: Remove dg-options for sh*-*-*.
--- ORIG/trunk/gcc/testsuite/gcc.c-torture/execute/pr44683.c2014-08-26
09:26:20.0
Mike Stump wrote:
>> BTW, Jeff's comment on pr39228.c makes me think that this test
>> can be moved into execute/ieee. Thought?
>
> I don’t see a down side, specially if we know that the test case is picky
> about ieee.
Thanks for suggestion. The patch below simply moves it into
execute/ieee.
Christian Bruel wrote:
> No regressions for sh-none-elf. OK for trunk ?
OK.
Regards,
kaz
Oleg Endo wrote:
> It seems that splitting multi-word insns before reload results in
> slightly better code on average (according to CSiBE). The attached
> patch implements that.
> Tested on rev. 204263 with
> make -k -j4 check RUNTESTFLAGS="--target_board=sh-sim
> \{-m2/-ml,-m2/-mb,-m2a/-mb,-m2a
Oleg Endo wrote:
> This adds a test case for PR 30807 which is based on the PR's attachment
> 17961. Tested with
> make -k check-gcc RUNTESTFLAGS="sh-torture.exp --target_board=sh-sim
> \{-m2/-ml,-m2/-mb,-m2a/-mb,-m4/-ml,-m4/-mb,-m4a/-ml,-m4a/-mb}
>
> OK to add?
OK with the changes suggested by
Oleg Endo wrote:
> The attached patch addresses one thing which I mentioned in PR 22553
> comment #27. Some of the FP insns take "arith_reg_operand", which seems
> a bit off. I haven't checked whether it fixes the original problem of
> PR 22553, but anyway it's probably better to use "fp_arith_r
arrange of barrier_align.
The attached patch is to fix this.
Tested on sh4-unknown-linux-gnu. Committed.
Regards,
kaz
--
2013-11-15 Kaz Kojima
* config/sh/sh.c (barrier_align): Return 0 when barrier_or_label
is null.
diff -up ORIG/trunk/gcc/config/sh/sh.c trunk/gcc/config/sh/sh.
Oleg Endo wrote:
> * config.gcc (SH extra_objs): Add sh_optimize_sett_clrt pass.o.
The usual way would be
* config.gcc (sh[123456789lbe]*-*-* | sh-*-*): Add
sh_optimize_sett_clrt.o to extra_objs.
OK with that change. Thanks!
Regards,
kaz
Richard Sandiford wrote:
> I have a patch to upgrade most genrecog warnings into errors. This patch
> fixes the ones for sh. There were some warnings about the match_operands
> in some define_peephole2s having constraints. There were also a few
> cases where a destination predicate allowed a co
Oleg Endo wrote:
> Currently GCC doesn't pass the --isa parameter to the assembler for SH
> targets other than SH2A and SH5. This makes the assembler accept any
> kind of ISA and happily produce e.g. SH2A code even though the target is
> SH4, which will then fail to link.
> The attach patch fixes
Oleg Endo wrote:
> This is a small SH cleanup patch. No functional changes -- at least not
> intended.
>
> The ROUND_ADVANCE macro can be replaced with CEIL from system.h, which
> makes it a bit clearer towards what something is rounded in the places
> where it's used.
>
> The ROUND_REG and PAS
Oleg Endo wrote:
> The attached patch is the same as posted in the PR as attachment 31283.
> In addition to the testing done by Kaz and Christian, I've also tested
> it with
> make -k check RUNTESTFLAGS="--target_board=sh-sim
> \{-m2/-ml,-m2/-mb,-m2a/-mb,-m4/-ml,-m4/-mb,-m4a/-ml,-m4a/-mb}"
>
> on
o fix it. Committed on trunk.
Regards,
kaz
--
2014-06-18 Kaz Kojima
PR target/61550
* config/sh/sh.c (prepare_move_operands): Don't process TLS
addresses here if reload in progress or completed.
--- ORIG/trunk/gcc/config/sh/sh.c 2014-06-17 21:21:32
Hi,
I've committed the attached patches to fix PR target/65979
which is a 5/6 regression in behalf of Oleg. The problematic
peephole can clobber a register before its use. See PR65979 for
details. The first patch is for trunk and the second one is
a slightly different version against gcc-5-bran
Uros Bizjak wrote:
> 2015-05-27 Uros Bizjak
>
> * config/linux/sh/futex_bits.h (sys_futex0) Change operands
> "op" and "val" to int.
>
> Untested.
>
> OK for mainline?
OK.
Although it looks obvious, I've confirmed that there is no build
issue/regression on sh4-unknown-linux-gnu wit
.
The patch disparages this case for RA. It fixes PR and various
libstdc++ test failures happen on trunk with -mlra. Tested on
sh4-unknown-linux-gnu with no new failures. Committed.
Regards,
kaz
--
2015-06-05 Kaz Kojima
PR target/66410
* config/sh/constraints.md (Sid
e above PR. Tested on sh4-unknown-linux-gnu.
Regards,
kaz
--
2015-05-04 Kaz Kojima
PR target/65987
* config/sh/sh.c (output_far_jump): Take into account crossing jumps.
(split_branches): Likewise.
diff --git a/config/sh/sh.c b/config/sh/sh.c
index 1cf6ed0..a4c9
Hi,
I've got odd ICEs with some bitmap operations when working on
sh-lra branch.
bitmap_ior and bitmap_ior_and_compl are the bitmap operations
for DST = A | B and DST = A | (B & ~KILL) respectively. It
looks they could make bitmaps with the wrong "current" pointer.
One small example is
DST: [(bi
Richard Biener wrote:
> Huh? Wasn't this fixed by Mike a few weeks ago?
Ugh. I've missed that thread. Sorry for the noise.
Regards,
kaz
Mike Stump wrote:
> While I had testing on my target, there were plenty of other regressions
> from trunk intermixed, so I wanted to do a normal bootstrap on Linux which
> I had not done yet. Kaz, you can check your patch, if the same, pretty
> sure it is, it has been approved and you can just ch
See
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=65153
for details.
Tested on sh4-unknwon-linux-gnu with no new failures. Committed
on trunk. I'll backport it to 4.9 branch later.
Regards,
kaz
--
2015-02-23 Kaz Kojima
PR target/65153
* config/sh/sh.md (movs
Oleg Endo wrote:
> These are the peepholes as mentioned in PR 65153 and in PR 61142. They
> try to wallpaper some bad RA choices and reduce the CSiBE code size by
> approx. 3.9K bytes.
>
> A problem I ran into with this one is that the peephole2 pass drops
> REG_INC notes, which makes the follow
, SH is still in
transition to LRA and we have no chance for 4.9.
Tested with sh4-unknown-linux-gnu on trunk and 4.9 branch.
Committed on trunk and 4.9.
Regards,
kaz
--
2015-03-03 Kaz Kojima
PR target/65249
* config/sh/sh.md (symGOT_load): Use R0 reg for operands[2] when
Oleg Endo wrote:
>> diff --git a/gcc/config/sh/t-linux b/gcc/config/sh/t-linux
>> index d33c638..22b8e3e 100644
>> --- a/gcc/config/sh/t-linux
>> +++ b/gcc/config/sh/t-linux
>> @@ -1,2 +1,3 @@
>> MULTILIB_DIRNAMES=
>> -MULTILIB_MATCHES =
>> +MULTILIB_MATCHES =
>> +MULTILIB_EXCEPTIONS = m2a m2a/
Oleg Endo wrote:
> Backporting of the fix went into GCC 5 seems to intrusive for the
> released branches. Hence I'd propose to remove the problematic patterns
> altogether, as they can silently generate wrong code. The attached
> patch does that. Tested briefly with 'make all' and with
> make -
Yoshinori Sato wrote:
>> * config/sh/t-linux (MULTILIB_EXCEPTIONS): Define for m2a cases.
[snip]
> It works fine.
Thanks for checking. I've committed it on trunk as revision 221287.
Regards,
kaz
Oleg Endo wrote:
> I think the same should be done in t-sh and in t-linux we could add the
> SH1 little endian exception, too ... just in case (see attached patch).
> I'm not sure whether TM_MULTILIB_EXCEPTIONS_CONFIG is needed in t-linux
> or not. Moreover, I don't understand why t-linux doesn't
Steven Bosscher wrote:
> This patch merely papers over another issue, probably a missing
> CROSSING_JUMP_P test.
Perhaps. Surely it has looked the current DBR is not so well for
crossing jumps and my fix might be a bit ad-hoc.
The first part of the patch could be rewritten with checking if one
o
n sh4-unknown-linux-gnu with no new failures for -mlra/-mno-lra.
There are no visible changes over generated objects with -mno-lra
in CSiBE. Committed on trunk.
Regards,
kaz
--
2015-06-10 Kaz Kojima
* config/sh/sh.md (tstsi_t): Add '?' modifier to 'r' alterna
1 - 100 of 397 matches
Mail list logo
