[i386] Do not omit the frame pointer at -O0

Ping for https://gcc.gnu.org/ml/gcc-patches/2018-06/msg01228.html Thanks in advance. -- Eric Botcazou

Re: [i386] Do not omit the frame pointer at -O0

> LGTM, but please note that the patch was already approved by Jeff on > 22th of June [1]. Sorry, I missed that... Thanks for pointing it out. -- Eric Botcazou

Fix old thinko in choose_multiplier

. Bootstrapped/regtested on x86-64/Linux, applied on the mainline as obvious. 2018-07-04 Eric Botcazou PR middle-end/86380 * expmed.c (choose_multiplier): Fix incorrect comparison with mask. -- Eric BotcazouIndex: expmed.c

Re: [PATCH, Ada] RISC-V: Initial riscv linux Ada port.

ures. Ada doesn't use trampolines if you define... > + Always_Compatible_Rep : constant Boolean := False; ...this to False. -- Eric Botcazou

Re: [PATCH, Ada] RISC-V: Initial riscv linux Ada port.

> Ada doesn't use trampolines if you define... > > > + Always_Compatible_Rep : constant Boolean := False; > > ...this to False. And also define TARGET_CUSTOM_FUNCTION_DESCRIPTORS for the architecture. -- Eric Botcazou

[Ada] Optimize calls to pure functions with by-ref In parameter

sense. Tested on x86-64/Linux, applied on the mainline. 2018-07-07 Eric Botcazou * gcc-interface/decl.c (gnat_to_gnu_param): Minor tweak. (gnat_to_gnu_subprog_type): New pure_flag local variable. Set it for a pure Ada function with a by-ref In parameter. Propagate it

[Ada] Do not generate debug info for actual subtypes

Eric Botcazou * gcc-interface/gigi.h (add_decl_expr): Adjust prototype. * gcc-interface/decl.c (gnat_to_gnu_entity): Remove useless test. * gcc-interface/trans.c (add_stmt_with_node): Remove exceptions. (add_decl_expr): Change type of second parameter and rename it

[Ada] Reduce -Wstack-usage false positives on variant records

This reduces the number of false positives of -Wstack-usage in the presence of variables whose nominal subtype is a discriminated record with a variant part. Tested on x86-64/Linux, applied on the mainline. 2018-07-07 Eric Botcazou * gcc-interface/decl.c (gnat_to_gnu_entity): Add

[c-family] Swich -fdump-ada-spec output for Ada 2012

/Linux, applied on the mainline. 2018-07-07 Eric Botcazou * c-ada-spec.c (to_ada_name): Remove index parameter. (pp_ada_tree_identifier): Likewise. (dump_ada_macros): Adjust call to to_ada_name. (struct overloaded_name_hash): Delete. (struct

Re: [PATCH, Ada] RISC-V: Initial riscv linux Ada port.

ected failures4 > # of expected failures 24 > # of unsupported tests 25 > > So yes, that solved my problem, and we have a working RISC-V Ada port > now. Thanks for the help. You're welcome. Are the 4 remaining failures related to stack checking? -- Eric Botcazou

Re: [PATCH, Ada] RISC-V: Initial riscv linux Ada port.

priate triplet to the list on line 29 when you have 15 seconds to kill. Patchlet preapproved. -- Eric Botcazou

[SPARC] Minor tweak

Tested on SPARC/Solaris, applied on the mainline. 2018-07-13 Eric Botcazou * config/sparc/sparc-protos.h (sparc_compute_frame_size): Delete. * config/sparc/sparc.c (sparc_compute_frame_size): Make static. -- Eric BotcazouIndex: config/sparc/sparc-protos.h

[testsuite] Minor tweak to 4 Aarch64 testcases

These 4 Aarch64 testcases use dg-xfail-if to disable themselves on ARM, while all the other equivalent testcases use dg-skip-if. The latter form is better because it doesn't unnecessarily pollute the testsuite report. Tested on arm-eabi, OK for the mainline? 2018-07-13 Eric Bot

[testsuite] Robustify target_tls_runtime check

As witnessed by the kludge added for MSP430 and Visium, the check doesn't really work (that's also visible for arm-eabi). Tested on x86_64-suse-linux, visium-elf and arm-eabi, OK for mainline? 2018-07-13 Eric Botcazou * lib/target-supports.exp (check_effective_target_t

Re: [testsuite] Minor tweak to 4 Aarch64 testcases

gt; should only be available on aarch64 and not arm. OK, thanks for the explanation. -- Eric Botcazou

Re: [PATCH, Ada] RISC-V: Initial riscv linux Ada port.

> I poked at this a little and noticed a difference between the x86_64 > support and the RISC-V support. The RISC-V C language port has char > as unsigned by default. The x86_64 port has char signed by default. > If I add a -fsigned-char option, then the testcase works as expected > for RISC-V.

Re: [PATCH] Properly unshare TYPE_SIZE_UNIT/DECL_FIELD_OFFSET (PR86216)

> It breaks Ada bootstrap. I guess Ada has variable-size types in > non-function scope (not sure how TYPE_SIZES_GIMPLIFIED works then > though). That said, r92495 moved the unshare_expr from layout_type > to gimplify_one_sizepos. See gimplify.c:763 and below. -- Eric Botcazou

[patch] Fix PR tree-optimization/86514

= _19; which is not equivalent. The proposed fix is to avoid bypassing a conversion to a boolean type from a type with greater precision in init_range_entry. Tested on x86_64-suse-linux, OK for the mainline? 2018-07-16 Eric Botcazou PR tree-optimization/86514

Re: [PATCH] Properly unshare TYPE_SIZE_UNIT/DECL_FIELD_OFFSET (PR86216)

types? walk_tree walks the DECL_EXPR of TYPE_DECL. I think that it's an old band-aid for types defined at library level in Ada. Let me experiment a bit with that. -- Eric Botcazou

[Ada] Fix spurious check failure for Character discriminant

position 127 & 128. It's also indirectly responsible for failure of gnat.dg/debug11.adb on RISC-V. Tested on x86-64/Linux, applied on the mainline, 8 and 7 branches. 2018-07-17 Eric Botcazou * gcc-interface/decl.c (choices_to_gnu): Rename parameters. Deal with an o

Re: [PATCH] enhance strlen to understand MEM_REF and partial overlaps (PR 86042, 86043)

| | Error detected around c52103c.adb:43:1 | -- Eric Botcazou

Re: [PATCH] enhance strlen to understand MEM_REF and partial overlaps (PR 86042, 86043)

ames of failing (and other > "interesting") tests. Why not just use "make mail-report.log" ? -- Eric Botcazou

Re: [PATCH] enhance strlen to understand MEM_REF and partial overlaps (PR 86042, 86043)

> What I was trying to highlight is that rolling my own solution like > this makes missing regressions more likely than having a shared > solution would. But 'make mail-report.log' is precisely the shared solution, no need to reinvent the wheel here. -- Eric Botcazou

Re: [patch] improve internals documentation for nested function descriptors

P-PA, so it's contradictory. Moreover I would really start with the "custom" case and not the standard case as was originally written, the "custom" case being the common case for targets. I'm not really convinced by the substitution misalignment -> tag either, but if others find the new version more understandable, then OK with me. -- Eric Botcazou

Re: [patch] improve internals documentation for nested function descriptors

what target-independent code does with a > descriptor once it has identified that's what it's got. Fair enough. -- Eric Botcazou

[Ada] Fix internal error on extension with interface at -O2

This is a regression present on the mainline, 10 and 9 branches, in the form of an internal error when a covariant-only thunk is inlined into its caller. Tested on x86-64/Linux, applied on the mainline, 10 and 9 branches. 2021-01-25 Eric Botcazou * gcc-interface/trans.c

[Ada] Fix PR ada/98228

nches. 2021-01-26 Eric Botcazou Marius Hillenbrand PR ada/98228 * gcc-interface/utils.c (maybe_pad_type): Test the size of the new packable type instead of its alignment for addressability's sake. -- Eric Botcazoudiff --git a/gcc/ada/gcc-interface/utils.c

Fix LTO bootstrap on Windows (PR lto/85574)

others don't. Bootstrapped in LTO mode on Windows, OK for all active branches? 2021-01-28 Eric Botcazou contrib/ PR lto/85574 * compare-lto: Deal with PE-COFF executables specifically. -- Eric Botcazoudiff --git a/contrib/compare-lto b/contrib/compare-lto index 17379e

Re: [PATCH] tree: Don't reuse types if TYPE_USER_ALIGN differ [PR94775]

> Bootstrapped/regtested on > * x86_64-pc-linux-gnu > * powerpc64le-unknown-linux-gnu > * aarch64-linux-gnu > ok for trunk? None of them is strict alignment though, isn't it? -- Eric Botcazou

Re: [PATCH] tree: Don't reuse types if TYPE_USER_ALIGN differ [PR94775]

le Farm: https://cfarm.tetaneutral.net/machines/list/ -- Eric Botcazou

Re: [PATCH] fold: Fix up strn{case,}cmp folding [PR98771]

in this scope 1777 | return build_int_cst (type, strncmp (p0, p1, MIN (s2, SIZE_MAX))); because /usr/include/stdint.h has: /* The ISO C99 standard specifies that in C++ implementations these macros should only be defined if explicitly requested. */ #if !defined __cplusplus || defined __STDC_LIMIT_MACROS -- Eric Botcazou

Re: [PATCH] fold: Fix up strn{case,}cmp folding [PR98771]

linek > > * fold-const-call.c (fold_const_call): Use INTTYPE_MAXIMUM (size_t) > instead of SIZE_MAX. Why not just: #ifndef SIZE_MAX # define SIZE_MAX INTTYPE_MAXIMUM (size_t) #endif just below UCHAR_MAX in system.h? -- Eric Botcazou

Re: [PATCH] fold: Fix up strn{case,}cmp folding [PR98771]

> Why not just: > > #ifndef SIZE_MAX > # define SIZE_MAX INTTYPE_MAXIMUM (size_t) > #endif > > just below UCHAR_MAX in system.h? Or rather just below #ifdef HAVE_STDINT_H #include #endif #ifdef HAVE_INTTYPES_H #include #endif -- Eric Botcazou

Re: [PATCH] fold: Fix up strn{case,}cmp folding [PR98771]

r macro. > Note, it now affects GCC 10 branch too. Yep, that's where it breaks things for us. :-) -- Eric Botcazou

Re: [PATCH] fold: Fix up strn{case,}cmp folding [PR98771]

> Whatever works, I can't test such patches except on Linux, so can you just > create a patch and test it on Solaris where it failed before? It fails on old Linux distros, e.g. RHES 5, not on Solaris as far as I know. -- Eric Botcazou

Re: [PATCH] fold: Fix up strn{case,}cmp folding [PR98771]

> Whatever works, I can't test such patches except on Linux, so can you just > create a patch and test it on Solaris where it failed before? Maybe a safer fix is the attached one. Tested on old RedHat and SuSE distros. * fold-const-call.c: Define __STDC_LIMIT_MACROS at the top. -- Eri

Re: [PATCH] fold: Fix up strn{case,}cmp folding [PR98771]

> But next time we use SIZE_MAX somewhere it is going to break again this way. > If we just define SIZE_MAX if not defined after all includes in system.h, I > think it is better than this. The existing practice seems to define the missing constants right after the corresponding include, e.g. MAP_

[Ada] Assorted LTO fixes

This polishes a few rough edges visible in LTO mode. Tested on x86-64/Linux, applied on mainline, 10 and 9 branches. 2021-02-03 Eric Botcazou * gcc-interface/decl.c (gnat_to_gnu_entity) : Make the two fields of the fat pointer type addressable, and do not make the

[Ada] Fix regression with partial rep clause on variant record type

It is present on the mainline, 10 and 9 branches, and can yield an incorrect layout when there is a partial representation clause on a discriminated record type with a variant part. Tested on x86-64/Linux, applied on mainline, 10 and 9 branches. 2021-02-03 Eric Botcazou * gcc

Fix PR rtl-optimization/96015

branch. 2021-02-09 Eric Botcazou PR rtl-optimization/96015 * reorg.c (skip_consecutive_labels): Minor comment tweaks. (relax_delay_slots): When deleting a jump to the next active instruction over a barrier, first delete the barrier if the jump is the only

[x86] Fix -freorder-blocks-and-partition glitch with Windows SEH

-functioning Ada compiler when you do a profiled bootstrap. Bootstrapped on x86-64/Windows, applied on all active branches as obvious. 2021-02-11 Eric Botcazou * config/i386/winnt.c (i386_pe_seh_unwind_emit): When switching to the cold section, emit a nop before the directive if the

Fix cast in df_worklist_dataflow_doublequeue

The existing cast to float gives weird results in the RTL dump files on x86 when the compiler is configured --with-fpmath=sse. Bootstrapped on x86/Linux, applied on the mainline as obvious. 2021-02-15 Eric Botcazou * df-core.c (df_worklist_dataflow_doublequeue): Use proper cast

Re: [PATCH 2/2] sparc: Run SUBTARGET_INIT_BUILTINS if it exists

2520 100644 > --- a/gcc/config/sparc/sparc.c > +++ b/gcc/config/sparc/sparc.c > @@ -10962,6 +10962,9 @@ sparc_init_builtins (void) > >if (TARGET_VIS) > sparc_vis_init_builtins (); > +#ifdef SUBTARGET_INIT_BUILTINS > + SUBTARGET_INIT_BUILTINS; > +#endif > } Missing blank line before the change. -- Eric Botcazou

[x86] Fix PR target/99264

sequence for SEH whatever the frame size. And it avoids a weird discrepancy between cases where the number of saved general regs is even and cases where it is odd. Tested on x86_64-w64-mingw32, OK for mainline, 10 and 9 branches? 2021-02-26 Eric Botcazou PR target/9926

[Ada] Fix PR ada/99095

This is a regression present on the mainline and 10 branch, where we fail to make the bounds explicit for the return value of a function returning an unconstrained array of a limited record type. Tested on x86-64/Linux, applied on mainline and 10 branch. 2021-03-02 Eric Botcazou PR

Re: [x86] Fix PR target/99264

0/9 branches as obvious. 2021-03-03 Eric Botcazou PR target/99234 * config/i386/i386.c (ix86_compute_frame_layout): For a SEH target, point back the hard frame pointer to its default location when the frame is larger than SEH_MAX_FRAME_SIZE. -- Eric Botcazoudiff --

Re: [PATCH] sparcv9: Disable -Wuninitialized warnings breaking bootstrap [PR92002]

t release branches do not have -Werror set, do they? -- Eric Botcazou

[patch] Fix PR rtl-optimization/99376

Hi, this is an undefined behavior spotted by the sanitizer that has managed to go unnoticed until now. Tested on x86-64/Linux, OK for the mainline? 2021-03-04 Eric Botcazou PR rtl-optimization/99376 * rtlanal.c (nonzero_bits1) : If the number of low-order zero bits

[Ada] Fix PR ada/99264

This fixes the build breakage introduced by the latest glibc release. Tested on x86-64/Linux, applied on mainline, 10 and 9 branches. 2021-03-05 Eric Botcazou PR ada/99264 * init.c (__gnat_alternate_sta) [Linux]: Remove preprocessor test on MINSIGSTKSZ and bump size

Re: [pushed] c++: Fix class NTTP constness handling [PR98810]

ot; for " dg-do 2 compile { target c++20 } " ERROR: g++.dg/cpp2a/nontype-class-defarg1.C -std=c++17: syntax error in target selector "target c++20" for " dg-do 2 compile { target c++20 } " On the contrary, the 10 branch is a in good shape. -- Eric Botcazou

Re: [pushed] c++: Fix class NTTP constness handling [PR98810]

> GCC 9 doesn't have the "c++20" target yet. > > * g++.dg/cpp2a/nontype-class-defarg1.C: Use target c++2a. Thanks! -- Eric Botcazou

Re: [PATCH v2] fix Ada bootstrap on Cygwin64 (PR bootstrap/94918)

mingw32.h" #endif at the appropriate spot in raise-gcc.c instead? -- Eric Botcazou

[patch] Fix PR C++/90448

a register. Bootstrapped/regtested on x86-64/Linux, PowerPC64/Linux and SPARC/Solaris, OK for the mainline and 10 branch? 2021-03-08 Eric Botcazou PR C++/90448 * calls.c (initialize_argument_information): When the argument is passed by reference, do not make a copy in

[patch] Fix PR fortran/96983

Hi, AFAICS the code in build_round_expr implicitly assumes that __float128 exists, which is *not* the common case among 64-bit architectures since "long double" is generally already 128-bit for them. Tested on x86-64/Linux and SPARC64/Linux, OK for the mainline? 2021-03-08 Eri

Re: [PATCH v2] fix Ada bootstrap on Cygwin64 (PR bootstrap/94918)

> This one worked. Is that what you had in mind? > > * raise-gcc.c: On Cygwin include mingw32.h to prevent > windows.h from including x86intrin.h or emmintrin.h. Yep, exactly, thanks, you may put it on whichever branch you need. -- Eric Botcazou

Re: [patch] Fix PR C++/90448

red about the issue for months, so please let's not the best be the enemy of the good... -- Eric Botcazou

Re: [PATCH 2/2] Ada: Remove debug line number for DECL_IGNORED_P functions

t; > 2021-07-24 Bernd Edlinger > > PR debug/101598 > * gcc-interface/trans.c (Subprogram_Body_to_gnu): Set the > DECL_SOURCE_LOCATION of DECL_IGNORED_P gnu_subprog_decl to > UNKNOWN_LOCATION. Is that really needed in DWARF 5? If no, I'm not sure that we want it. -- Eric Botcazou

Re: [PATCH 2/2] Ada: Remove debug line number for DECL_IGNORED_P functions

But we still want to have location information in the compiler itself to debug it. -- Eric Botcazou

Fix PR tree-optimization/101626

This is a regression present on the mainline, caused by an oversight of mine in an earlier fix for SRA, whereby I forgot to exclude cases for reverse SSO. Tested on x86-64/Linux, applied on the mainline as obvious. 2021-08-05 Eric Botcazuo PR tree-optimization/101626 * tree-

Re: [PATCH] Fix ICE when mixing VLAs and statement expressions [PR91038]

If the comment is right, then Martin's patch should break Ada indeed. -- Eric Botcazou

Re: [PATCH 2/2] Ada: Remove debug line number for DECL_IGNORED_P functions

the location for Test2 is better > than no debug info at all? But we want no debug info at all for these IP functions. > What do you think? I guess I still don't understand why DECL_IGNORED_P was changed. -- Eric Botcazou

Re: [PATCH 2/2] Ada: Remove debug line number for DECL_IGNORED_P functions

guess Bernd's patch is acceptable, modulo a small tweak for -gnatD. Let me experiment a little bit though. -- Eric Botcazou

Re: [PATCH] Fix ICE when mixing VLAs and statement expressions [PR91038]

use-linux) GCC error:| | in gimplify_expr, at gimplify.c:15019| | Error detected around vect1.adb:6:23 probably because array_ref_low_bound, where it is normally eliminated, cannot do its jub properly when it is called after the base expression is gimplified. -- Eric Botcazou

Re: [PATCH 2/2] Ada: Remove debug line number for DECL_IGNORED_P functions

# DEBUG x => x_1(D) > _4 = x_1(D); > # DEBUG x => NULL > retval.5_2 = _4; > return retval.5_2; > > } > > > the line numbers are gone, and the function has DECL_IGNORED_P, > but still a useful DECL_SOURCE_LOCATION, I don't know > where the DECL_SOURCE_LOCATION can be seen in the dump files, > but from debugging this effect, I know that quite well. > > This second effect is why as a special case DECL_IGNORED_P functions > with valid looking DECL_SOURCE_LOCATION have now a .loc statement, > because that is less surprising to a user than having no line numbers > at all here. OK, thanks for the explanation, the patch is OK then. -- Eric Botcazou

Re: [PATCH][v2] Adjust volatile handling of the operand scanner

What do we do for other similar flags, e.g. TREE_READONLY? -- Eric Botcazou

Re: [PATCH][v2] Adjust volatile handling of the operand scanner

when the FIELD_DECL has > TREE_THIS_VOLATILE set. This would be weird semantics in my opinion. > I guess I'll do one more experiment and add verification that > TREE_THIS_VOLATILE on COMPONENT_REFs and FIELD_DECLs is consistent > and see where that trips. Sounds good to me. -- Eric Botcazou

Re: [PATCH][v2] Adjust volatile handling of the operand scanner

> So I'm leaning towards leaving build3 alone and fixing up frontends > as issues pop up. FWIW fine with me. -- Eric Botcazou

Small tweak to expand_used_vars

This completes the replacement of DECL_ATTRIBUTES (current_function_decl) with the attribs local variable. Tested on x86-64/Linux, applied on the mainline as obvious. 2021-08-11 Eric Botcazuo * cfgexpand.c (expand_used_vars): Reuse attribs local variable. -- Eric Botcazoudiff --gi

[patch] Make -no-pie option work for native Windows

compiler. Bootstrapped on x86 and x86-64/Windows, w/ and w/o binutils 2.36, OK for the mainline and 11 branch? 2021-08-11 Eric Botcazou * configure.ac (PE linker --disable-dynamicbase support): New check. * configure: Regenerate. * config.in: Likewise

Re: [patch] Make -no-pie option work for native Windows

> Looks good to me. Do you have push permissions? Thanks. Yes, see the preceding message on gcc-patches@, so applied. -- Eric Botcazou

Disable GNAT encodings by default

Eric Botcazou * dwarf2out.c (add_data_member_location_attribute): Use GNAT encodings only when -fgnat-encodings=all is specified. (add_bound_info): Likewise. (add_byte_size_attribute): Likewise. (gen_member_die): Likewise. -- Eric Botcazoudiff --git a

[patch] Fix regression in debug info for Ada with DWARF 5

mainline and 11 branch? 2021-08-16 Eric Botcazou * dwarf2out.c (add_scalar_info): Deal with DW_AT_data_bit_offset. -- Eric Botcazoudiff --git a/gcc/dwarf2out.c b/gcc/dwarf2out.c index 4bcd3313fee..ba0a6d6ed60 100644 --- a/gcc/dwarf2out.c +++ b/gcc/dwarf2out.c @@ -21253,6 +21253,7

Re: [PATCH] Make sure -fexceptions is enabled when -fnon-call-exceptions is

> This makes -fexceptions enabled by -fnon-call-exceptions, removing > the odd state of !flag_exceptions && flag_non_call_exceptions from > middle-end consideration. FWIW fine with me. -- Eric Botcazou

[Ada] Add preliminary support for 128-bit integer types

This is only the gigi part, in preparation for the bulk of the implementation. Tested on x86_64-suse-linux, applied on the mainline. 2020-09-12 Eric Botcazou * gcc-interface/gigi.h (standard_datatypes): Add ADT_mulv128_decl. (mulv128_decl): New macro

[Ada] Minor tweak to line debug info

This prevents the SLOC of the expression for a tag from being present in the line debug info every time it is referenced for coverage purposes. Tested on x86_64-suse-linux, applied on the mainline. 2020-09-12 Eric Botcazou * gcc-interface/trans.c (gnat_to_gnu) : Clear the

[Ada] Accept absolute address clause for array of UNC nominal subtype

This changes the compiler to accept again absolute address clause for aliased array of unconstrained nominal subtype, instead of erroring out in this case. Tested on x86_64-suse-linux, applied on the mainline. 2020-09-12 Eric Botcazou * gcc-interface/decl.c (gnat_to_gnu_entity

[Ada] Fix small inconsistency in new predicate

This can result on the mainline in a segfault when an object declared at library level is used in the declaration of another, local object. Tested on x86_64-suse-linux, applied on the mainline. 2020-09-12 Eric Botcazou * gcc-interface/trans.c (lvalue_for_aggr_p) : Return

[patch] Fix dangling references in thunks at -O0

to its target, so it cannot use a tail call to invoke it. Tested on x86_64-suse-linux, OK for the mainline? 2020-09-14 Eric Botcazou * cgraphunit.c (cgraph_node::expand_thunk): Force a tail call only when optimizing. 2020-09-14 Eric Botcazou * gnat.dg

Re: [patch] Fix dangling references in thunks at -O0

do we, besides from this very single spot, simply never tailcall at -O0 > and thus never hit this latent issue? Presumably yes, tail calling is an optimization like the others. > How does this change the debug experience at -O0 when GIMPLE thunks > are used? In Ada this doesn't change much since thunks have line debug info. -- Eric Botcazou

Re: [patch] Fix dangling references in thunks at -O0

t -O0. I guess tail-calling is > forced here because gimple analysis sometimes would not mark the call. The low-level assembly thunks simply support the most simple thunks, see the condition in expand_thunk. Moreover targets can opt out as they wish. -- Eric Botcazou

Re: [patch] Fix dangling references in thunks at -O0

> That being said, I can add another bit to cfun, e.g. is_gimple_thunk, and > force the use of a register above only in case the bit is true. In fact I can probably reuse cfun->tail_call_marked for this purpose. -- Eric Botcazou

Re: [patch] Fix dangling references in thunks at -O0

> In fact I can probably reuse cfun->tail_call_marked for this purpose. Like so. * cgraphunit.c (cgraph_node::expand_thunk): Make sure to set cfun->tail_call_marked when forcing a tail call. * function.c (assign_parm_setup_reg): Always use a register to retrieve a

[patch] Fix pessimization in EH cleanup pass

ne debug info to avoid a problematic inheritance for coverage measurement. Tested on x86_64-suse-linux, OK for the mainline? 2020-09-15 Eric Botcazou Pierre-Marie Derodat * tree-eh.c (lower_try_finally_dup_block): Backward propagate slocs to stack restore built

Re: [patch] Fix pessimization in EH cleanup pass

> So it breaks PR93199 again? Indeed, although there is no regression in the testsuite AFAICS. I guess that we can do the new walk before and not instead of the post-order traversal. Revised patch attached, same ChangeLog. -- Eric Botcazoudiff --git a/gcc/tree-eh.c b/gcc/tree-eh.c index 4246d

Re: [patch] Fix dangling references in thunks at -O0

> This introduces an ICE building the glibc testsuite for alpha (bisected), > s390 and sparc (symptoms appear the same, not bisected to confirm the > exact revision). See bug 97078. Fixed thus, tested on x86_64-suse-linux, applied on mainline as obvious. PR middle-end/97078 * fu

Re: [PATCH] optabs: Don't reuse target for multi-word expansions if it overlaps operand(s) [PR97073]

ntioned_p between target > and operand(s) and if it returns true, force a pseudo as target. > > * gcc.c-torture/execute/pr97073.c: New test. This looks good to me. -- Eric Botcazou

[Ada] Fix bogus alignment warning on address clause

This is a regression present on the mainline and 10 branch: the compiler gives a bogus alignment warning on an address clause and a discriminated record type with variable size. Tested on x86_64-suse-linux, applied on the mainline and 10 branch. 2020-09-28 Eric Botcazou * gcc

[Ada] Add missing end location information

: https://gcc.gnu.org/pipermail/gcc-patches/2020-September/553951.html so I reverted it. 2020-09-28 Eric Botcazou * gcc-interface/trans.c (Subprogram_Body_to_gnu): Set the end locus of body and declaration earlier. -- Eric Botcazoudiff --git a/gcc/ada/gcc-interface/trans.c

[patch] Do not use doloop pattern with pragma Unroll

unrolling pass to compute the best number of iterations. Tested on PowerPC64/Linux and x86-64/Linux, OK for the mainline? 2020-09-28 Eric Botcazou * tree-ssa-loop-ivopts.c (analyze_and_mark_doloop_use): Bail out if the loop is subject to a pragma Unroll with no spec

[rs6000] Avoid useless masking of count operand for rotation

7;s done by the hardware so it would be nice to get rid of it. I have attached a couple of patches to that effect: the first one adds new instructions while the second one only adds splitters. Tested on PowerPC64/Linux, OK (which one) for the mainline? 2020-09-29 Eric Botcazou

Re: [rs6000] Avoid useless masking of count operand for rotation

this won't be handled before combine (or what do I miss?), it is > fine to do splitters only (splitters for combine). But the other > approach is fine as well. Patch #2 uses define_and_split like the x86 back-end; moreover, we already have define_and_split for the dot variants so maybe it's the best way to go? -- Eric Botcazou

[patch] Fix PR tree-optimization/92131

combining ranges with symbolic expressions. Tested on x86_64-suse-linux, OK for all active branches? 2019-10-23 Eric Botcazou PR tree-optimization/92131 * tree-vrp.c (extract_range_from_plus_minus_expr): If the resulting range would be symbolic, drop to varying for any

Re: [RFA][1/3] Remove Cell Broadband Engine SPU targets

> OK for all 3 patches. Similarly for removing any other remnants you > might find later. I have removed left-overs in htdocs/backends.html -- Eric Botcazoudiff --git a/htdocs/backends.html b/htdocs/backends.html index 6e212b24..c9449065 100644 --- a/htdocs/backends.html +++ b/htdocs/backends.h

Fix PR testsuite/92302

This makes the test more robust by relaxing the regexp. Tested on SPARC/Solaris 11, applied on the mainline. 2019-11-04 Eric Botcazou PR testsuite/92302 * gcc.target/sparc/sparc-ret-3.c: Accept more registers in address. -- Eric BotcazouIndex: gcc.target/sparc/sparc-ret-3

[SPARC] Fix PR target/92095

SPARC port, which turns out to be more disruptive than initially thought. Bootstrapped/regtested on SPARC64/Linux and SPARC/Solaris, applied on mainline and 9/8 branches. 2019-11-08 Eric Botcazou PR target/92095 * config/sparc/sparc-protos.h (output_load_pcrel_sym): Declare

Re: [PATCH, GCC] Fix unrolling check.

that shows up as excess error due -fopt-info. This > option was added in r202077 but I am not sure why this particular test > was chosen for it. That's a regression, there should be no unrolling. -- Eric Botcazou

Re: [PATCH, GCC] Fix unrolling check.

pragma and compare, there should be no changes. > What "do we always do"? What we do in the absence of specific unrolling directives for the loop. -- Eric Botcazou

Re: Fix ICE in ipa-cp when mixing -O0 and -O2 code in LTO

ugger set breakpoints and inspect subprogram parameters on exception # related events. We could probably try wih -Og these days. -- Eric Botcazou

[Ada] Fix -fdump-ada-spec issue with array field

This is a regression present on mainline and 9 branch: if the component type of array type is a structure declared in another file, then the binding would contain a local declaration for this structure type. Tested on x86_64-suse-linux, applied on the mainline and 9 branch. 2019-11-13 Eric

Re: [pushed] Ada : Fix bootstrap after r11-4793.

L_IS_UNDECLARED_BUILTIN. Thanks for fixing this! -- Eric Botcazou

  1   2   3   4   5   6   7   8   9   10   >