Re: [SFN] Bootstrap broken

2017-12-19 Thread Alexandre Oliva
On Dec 19, 2017, Jakub Jelinek wrote: > Has the patch you've posted passed bootstrap/regtest? Yeah, here's the latest version, that survived O1, O2 and O3 bootstraps with bootstrap-debug (-g0 for stage2), bootstrap-debug-lean (-fcompare-debug for stage3) and

Re: [PATCH][i386,AVX] Enable VPCLMULQDQ support

2017-12-19 Thread Kirill Yukhin
Hello Julia, On 09 Nov 09:47, Koval, Julia wrote: > Hi, this patch enables VPCLMULQDQ instruction from VPCLMULQDQ isaset, defined > here: > https://software.intel.com/sites/default/files/managed/c5/15/architecture-instruction-set-extensions-programming-reference.pdf > > Ok for trunk? Your patch

Re: [PATCH] Fix PR83452

2017-12-19 Thread Alan Modra
On Tue, Dec 19, 2017 at 03:37:10PM +0100, Richard Biener wrote: > + while ((gnu_lto = memchr (gnu_lto, 'g', > + strings + strsz - gnu_lto))) > + if (strncmp (gnu_lto, "gnu_lto_v1", > + strings + strsz - gnu_lto) ==

Re: [PATCH] skip memcpy-6.c for failing targets (PR 83483)

2017-12-19 Thread Jeff Law
On 12/19/2017 05:18 PM, Martin Sebor wrote: > As discussed in PR 83483, the gcc.dg/memcpy-6.c test I added > along with the -Wrestrict changes fails on many targets because > of bad assumptions on my part.  The test wasn't quite ready to > commit but I forgot to get back to it and finish it up. >

Re: [PATCH] Eliminate location wrappers in tree_nop_conversion/STRIP_NOPS

2017-12-19 Thread Jeff Law
On 12/19/2017 05:44 PM, David Malcolm wrote: > On Tue, 2017-12-19 at 23:03 +0100, Jakub Jelinek wrote: >> On Tue, Dec 19, 2017 at 04:59:34PM -0500, Jason Merrill wrote: Or do you mean it should strip just the special VIEW_CONVERT_EXPR that has type identical to the operand's type? >>>

Re: [v2 of PATCH 15/14] Use fold_for_warn in get_atomic_generic_size

2017-12-19 Thread Jason Merrill
On Tue, Dec 19, 2017 at 7:53 PM, David Malcolm wrote: > On Tue, 2017-12-19 at 15:35 -0500, Jason Merrill wrote: >> On Sat, Dec 16, 2017 at 8:12 PM, David Malcolm >> wrote: >> > I rebased the v2 patchkit; here's an extra patch to fix an issue >> > with it

Re: [PATCH] Fix PR83418

2017-12-19 Thread Jeff Law
On 12/15/2017 09:30 AM, Richard Biener wrote: > On December 15, 2017 5:27:14 PM GMT+01:00, Jeff Law wrote: >> On 12/15/2017 01:10 AM, Richard Biener wrote: >>> On Thu, 14 Dec 2017, Richard Biener wrote: >>> On December 14, 2017 4:43:42 PM GMT+01:00, Jeff Law

[v2 of PATCH 15/14] Use fold_for_warn in get_atomic_generic_size

2017-12-19 Thread David Malcolm
On Tue, 2017-12-19 at 15:35 -0500, Jason Merrill wrote: > On Sat, Dec 16, 2017 at 8:12 PM, David Malcolm > wrote: > > I rebased the v2 patchkit; here's an extra patch to fix an issue > > with it uncovered by a recently-added testcase (in r254990). > > > > With the patch kit,

[PATCH] Eliminate location wrappers in tree_nop_conversion/STRIP_NOPS

2017-12-19 Thread David Malcolm
On Tue, 2017-12-19 at 23:03 +0100, Jakub Jelinek wrote: > On Tue, Dec 19, 2017 at 04:59:34PM -0500, Jason Merrill wrote: > > > Or do you mean it should strip just the special VIEW_CONVERT_EXPR > > > that has type identical to the operand's type? > > > > That; interpreting something as the same

Re: [PATCH,PTX] Add support for CUDA 9

2017-12-19 Thread Tom de Vries
On 12/20/2017 12:25 AM, Cesar Philippidis wrote: In CUDA 9, Nvidia removed support for treating the labels of functions as generic address spaces as part of their PTX 6.0 changes. More specifically,

Re: [PATCH] Improve alloca alignment

2017-12-19 Thread Jeff Law
On 12/13/2017 04:17 PM, Eric Botcazou wrote: >> No clue, but ISTM that it should. Eric, can you try that and see if it >> addresses these problems? I'd really like to get this wrapped up, but I >> don't have access to any sparc systems to test it myself. > > Yes, the INIT_EXPANDERS trick works

Re: [14/nn] Add helpers for shift count modes

2017-12-19 Thread Jeff Law
On 12/19/2017 12:13 PM, Richard Sandiford wrote: > Richard Sandiford writes: >> Richard Biener writes: >>> On Fri, Dec 15, 2017 at 1:48 AM, Richard Sandiford >>> wrote: Richard Biener

Re: [PATCH, PR83492] Fix selection of aarch64 big-endian shift parameters based on __AARCH64EB__

2017-12-19 Thread Jeff Law
On 12/19/2017 11:23 AM, Jakub Jelinek wrote: > On Tue, Dec 19, 2017 at 07:01:44PM +0100, Michael Weiser wrote: >> Hi Kyrill, >> >> On Tue, Dec 19, 2017 at 05:35:03PM +, Kyrill Tkachov wrote: >> below patch fixes PR 83492. >>> I agree with your analysis of the bug and your patch will fix

Re: [PATCH] make canonicalize_condition keep its promise

2017-12-19 Thread Segher Boessenkool
On Tue, Dec 19, 2017 at 04:56:16PM -0700, Jeff Law wrote: > With that in mind your patch is fine. Thanks for the reviewing :-) > I will note that I find it highly confusing that we attach a simple loop > descriptor to a loop that is not a simple loop. But clearly you didn't > introduce that

[PATCH] skip memcpy-6.c for failing targets (PR 83483)

2017-12-19 Thread Martin Sebor
As discussed in PR 83483, the gcc.dg/memcpy-6.c test I added along with the -Wrestrict changes fails on many targets because of bad assumptions on my part. The test wasn't quite ready to commit but I forgot to get back to it and finish it up. Rather than remove it, the attached patch skips

Re: [PATCH] make canonicalize_condition keep its promise

2017-12-19 Thread Segher Boessenkool
Hi Jeff, On Tue, Dec 19, 2017 at 04:40:23PM -0700, Jeff Law wrote: > On 12/15/2017 02:16 AM, Segher Boessenkool wrote: > >> The only way to get into check_simple_exit is via find_simple_exit which > >> is only called from get_simple_loop_desc. > >> > >> And if you're calling get_simple_loop_desc,

Re: [PATCH, PR83423] Don't call targetm.calls.static_chain in non-static function

2017-12-19 Thread Jeff Law
On 12/19/2017 09:34 AM, Tom de Vries wrote: > Hi, > > this patch fixes PR83423. > > > The default_static_chain hook has implemented a sorry if both > STATIC_CHAIN_INCOMING_REGNUM and STATIC_CHAIN_REGNUM are undefined: > ... > rtx > default_static_chain (const_tree ARG_UNUSED (fndecl_or_type),

Re: [PATCH] Fix RTL DCE ICEs due to TYPE_EMPTY_P (PR c++/83490)

2017-12-19 Thread Jeff Law
On 12/19/2017 03:02 PM, Jakub Jelinek wrote: > Hi! > > The following testcase ICEs because we get: > (call_insn/u 7 4 14 2 (set (reg:QI 0 ax) > (call (mem:QI (symbol_ref:DI ("_Zlt1AS_") [flags 0x3] 0x7efd00f7fc00 operator<>) [0 operator< S1 A8]) > (const_int 0 [0]))) 690

Re: [PATCH] make canonicalize_condition keep its promise

2017-12-19 Thread Jeff Law
On 12/14/2017 03:22 PM, Aaron Sawdey wrote: >> >> Understood.  But what I still struggle with is how you're getting >> into >> check_simple_exit to begin with and whether or not that should be >> happening. >> >> >> The only way to get into check_simple_exit is via find_simple_exit >> which >> is

Re: [PATCH] make canonicalize_condition keep its promise

2017-12-19 Thread Jeff Law
On 12/15/2017 02:16 AM, Segher Boessenkool wrote: > On Thu, Dec 14, 2017 at 01:43:35PM -0700, Jeff Law wrote: >> On 11/21/2017 10:45 AM, Aaron Sawdey wrote: >>> There is no existing loop structure. This starts with a memcmp() call >>> and then goes down through the builtin expansion mechanism,

[PATCH,PTX] Add support for CUDA 9

2017-12-19 Thread Cesar Philippidis
In CUDA 9, Nvidia removed support for treating the labels of functions as generic address spaces as part of their PTX 6.0 changes. More specifically, : Support for taking address of labels, using

Re: [v2 of PATCH 06/14] Strip location wrappers in operand_equal_p

2017-12-19 Thread Jakub Jelinek
On Tue, Dec 19, 2017 at 04:59:34PM -0500, Jason Merrill wrote: > > Or do you mean it should strip just the special VIEW_CONVERT_EXPR > > that has type identical to the operand's type? > > That; interpreting something as the same type seems like a nop. Ok, that makes sense. Jakub

[PATCH] Fix RTL DCE ICEs due to TYPE_EMPTY_P (PR c++/83490)

2017-12-19 Thread Jakub Jelinek
Hi! The following testcase ICEs because we get: (call_insn/u 7 4 14 2 (set (reg:QI 0 ax) (call (mem:QI (symbol_ref:DI ("_Zlt1AS_") [flags 0x3] ) [0 operator< S1 A8]) (const_int 0 [0]))) 690 {*call_value} (expr_list:REG_EH_REGION (const_int 0 [0]) (nil))

Re: [v2 of PATCH 06/14] Strip location wrappers in operand_equal_p

2017-12-19 Thread Jason Merrill
On Tue, Dec 19, 2017 at 3:49 PM, Jakub Jelinek wrote: > On Tue, Dec 19, 2017 at 03:13:13PM -0500, Jason Merrill wrote: >> On 12/17/2017 09:07 PM, David Malcolm wrote: >> > On Mon, 2017-12-11 at 18:37 -0500, Jason Merrill wrote: >> > > On 11/10/2017 04:45 PM, David Malcolm wrote:

Re: [PR59319] output friends in debug info

2017-12-19 Thread Alexandre Oliva
On Dec 14, 2017, Jason Merrill wrote: > On 12/07/2017 04:04 PM, Alexandre Oliva wrote: >> For other templates, I ended up writing code to look for >> specializations in the hashtables of decl or type specializations. >> That's not exactly efficient, but it gets the job done.

Re: [PATCH][arm] PR target/82975: Guard against reg_renumber being NULL in arm.h

2017-12-19 Thread Jakub Jelinek
On Tue, Dec 19, 2017 at 04:58:37PM +, Kyrill Tkachov wrote: > 2017-12-19 Kyrylo Tkachov > > PR target/82975 > * gcc.dg/pr82975.c: New test. The testcase FAILs on x86_64-linux/i686-linux and probably on all non-arm/aarch64 targets. Fixed thusly, committed

Fix Debug DR2354

2017-12-19 Thread François Dumont
Hi   I plan to apply attached patch tomorrow to fix those _GLIBCXX_DEBUG failures: FAIL: 23_containers/map/modifiers/insert/dr2354.cc (test for excess errors) FAIL: 23_containers/multimap/modifiers/insert/dr2354.cc (test for excess errors) FAIL: 23_containers/unordered_map/insert/dr2354.cc

Re: [PATCH] Small load strlen opt follow-up improvement (PR tree-optimization/83444)

2017-12-19 Thread Jeff Law
On 12/19/2017 10:23 AM, Jakub Jelinek wrote: > Hi! > > As the testcase shows, we weren't handling the case where we have > _34 = MEM_REF[, 4B]; > - we would (if it exists) find the string length for b.a, but not > for the second field. Using get_addr_stridx handles that well. > The reason to use

Re: [v2 of PATCH 06/14] Strip location wrappers in operand_equal_p

2017-12-19 Thread Jakub Jelinek
On Tue, Dec 19, 2017 at 03:13:13PM -0500, Jason Merrill wrote: > On 12/17/2017 09:07 PM, David Malcolm wrote: > > On Mon, 2017-12-11 at 18:37 -0500, Jason Merrill wrote: > > > On 11/10/2017 04:45 PM, David Malcolm wrote: > > > > gcc/c-family/ChangeLog: > > > > * c-warn.c

Re: [C++ PING] Re: [C++ Patch, V2] PR 82593 ("Internal compiler error: in process_init_constructor_array, at cp/typeck2.c:1294")

2017-12-19 Thread Jason Merrill
On 11/30/2017 01:19 PM, Paolo Carlini wrote: On 17/11/2017 15:09, Paolo Carlini wrote: Hi again, I managed to spend much more time on the issue and I'm starting a new thread with a mature - IMHO - proposal: the big thing is the use of the existing check_array_designated_initializer in

Re: [12/13] Use ssizetype selectors for autovectorised VEC_PERM_EXPRs

2017-12-19 Thread Richard Sandiford
Ping Richard Sandiford writes: > The previous patches mean that there's no reason that constant > VEC_PERM_EXPRs need to have the same shape as the data inputs. > This patch makes the autovectoriser use ssizetype elements instead, > so that indices don't get

Re: [13/13] [AArch64] Use vec_perm_indices helper routines

2017-12-19 Thread Richard Sandiford
Ping Richard Sandiford writes: > This patch makes the AArch64 vec_perm_const code use the new > vec_perm_indices routines, instead of checking each element individually. > This means that they extend naturally to variable-length vectors. > > Also, aarch64_evpc_dup

Re: [10/13] Rework VEC_PERM_EXPR folding

2017-12-19 Thread Richard Sandiford
Ping Richard Sandiford writes: > This patch reworks the VEC_PERM_EXPR folding so that more of it works > for variable-length vectors. E.g. it means that we can now recognise > variable-length permutes that reduce to a single vector, or cases in > which a

Re: [11/13] Use vec_perm_builder::series_p in shift_amt_for_vec_perm_mask

2017-12-19 Thread Richard Sandiford
Ping Richard Sandiford writes: > This patch makes shift_amt_for_vec_perm_mask use series_p to check > for the simple case of a natural linear series before falling back > to testing each element individually. The series_p test works with > variable-length vectors

Re: [09/13] Use explicit encodings for simple permutes

2017-12-19 Thread Richard Sandiford
Ping Richard Sandiford writes: > This patch makes users of vec_perm_builders use the compressed encoding > where possible. This means that they work with variable-length vectors. > > > 2017-12-09 Richard Sandiford > > gcc/ > *

Re: [PATCH 15/14] Use fold_for_warn in get_atomic_generic_size

2017-12-19 Thread Jason Merrill
On Sat, Dec 16, 2017 at 8:12 PM, David Malcolm wrote: > I rebased the v2 patchkit; here's an extra patch to fix an issue > with it uncovered by a recently-added testcase (in r254990). > > With the patch kit, but without this patch, g++'s > c-c++-common/pr83059.c > fails to

Re: [PATCH] Fix up one mishandled plural in diagnostics in gimple-ssa-sprintf.c

2017-12-19 Thread Jeff Law
On 12/19/2017 10:20 AM, Jakub Jelinek wrote: > Hi! > > This inform is correct in english, but wrong in many other languages. > > Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk? > > 2017-12-19 Jakub Jelinek > > * gimple-ssa-sprintf.c

[committed][PR tree-optimization/83477] Always record a result range for PHIs

2017-12-19 Thread Jeff Law
BZ83477 is another case where we had an VR_UNDEFINED range for source operand during jump threading. In this instance we had a PHI argument that was a constant. During threading we record that the PHI result has the constant as its value. I mistakenly thought that would avoid the need to mess

Re: [v2 of PATCH 06/14] Strip location wrappers in operand_equal_p

2017-12-19 Thread Jason Merrill
On 12/17/2017 09:07 PM, David Malcolm wrote: On Mon, 2017-12-11 at 18:37 -0500, Jason Merrill wrote: On 11/10/2017 04:45 PM, David Malcolm wrote: gcc/c-family/ChangeLog: * c-warn.c (sizeof_pointer_memaccess_warning): Strip any location wrappers from src and dest. Here the

Re: [PATCH] C++: handle locations wrappers when calling warn_for_memset

2017-12-19 Thread Jason Merrill
On 12/16/2017 03:01 PM, Martin Sebor wrote: On 12/16/2017 06:12 AM, David Malcolm wrote: On Mon, 2017-12-11 at 18:36 -0500, Jason Merrill wrote: On 11/10/2017 04:45 PM, David Malcolm wrote: We need to strip away location wrappers in tree.c predicates like integer_zerop, otherwise they fail

Re: [PATCH] rl78 anddi3 improvement

2017-12-19 Thread DJ Delorie
Right, when doing 64-bit operations on an 8-bit mcu with limited registers, a hand-written assembler routine that you call as needed will beat anything gcc spits out - for size-per-call. And I had a lot of trouble getting gcc to deal with the rl78's limited register set and addressing modes -

Re: [v2 of PATCH 13/14] c-format.c: handle location wrappers

2017-12-19 Thread Jason Merrill
On 12/17/2017 11:29 AM, David Malcolm wrote: On Mon, 2017-12-11 at 18:45 -0500, Jason Merrill wrote: On 11/10/2017 04:45 PM, David Malcolm wrote: + STRIP_ANY_LOCATION_WRAPPER (format_tree); + if (VAR_P (format_tree)) { /* Pull out a constant value if the front end

[PATCH] Add NON_RECURSIVE attribute for procedures

2017-12-19 Thread Janne Blomqvist
Fortran 2018 makes procedures recursive by default (effectively making the existing RECURSIVE attribute a no-op). Instead it adds a NON_RECURSIVE attribute that a programmer can use to mark a procedure that may not be called recursively. This patch adds support for that. Regtested on

Re: C++ PATCH to add a test for c++/82231

2017-12-19 Thread Jason Merrill
OK. On Tue, Dec 19, 2017 at 6:56 AM, Marek Polacek wrote: > This bug was fixed in r255430, but the test looks different enough so we > better > add it. > > Tested on x86_64-linux, ok for trunk? > > 2017-12-19 Marek Polacek > > PR c++/82231 >

Re: [SFN] Bootstrap broken

2017-12-19 Thread Jakub Jelinek
On Tue, Dec 19, 2017 at 04:29:53PM -0200, Alexandre Oliva wrote: > On Dec 15, 2017, Jakub Jelinek wrote: > > > We want to verify there are no statements before labels. > > Erhm, actually... > > We already verify that in gimple_verify_flow_info. > > Is there any point in

Re: [SFN] Bootstrap broken

2017-12-19 Thread Alexandre Oliva
On Dec 15, 2017, Jakub Jelinek wrote: > We want to verify there are no statements before labels. Erhm, actually... We already verify that in gimple_verify_flow_info. Is there any point in having such redundant (?) verification? -- Alexandre Oliva, freedom fighter

Re: [14/nn] Add helpers for shift count modes

2017-12-19 Thread Richard Sandiford
Richard Sandiford writes: > Richard Biener writes: >> On Fri, Dec 15, 2017 at 1:48 AM, Richard Sandiford >> wrote: >>> Richard Biener writes: On Mon, Nov 20, 2017 at 10:02

Re: [PATCH] Fix up one mishandled plural in diagnostics in gimple-ssa-sprintf.c

2017-12-19 Thread Martin Sebor
On 12/19/2017 11:28 AM, Jakub Jelinek wrote: On Tue, Dec 19, 2017 at 11:19:44AM -0700, Martin Sebor wrote: On 12/19/2017 10:58 AM, Jakub Jelinek wrote: On Tue, Dec 19, 2017 at 10:49:07AM -0700, Martin Sebor wrote: Can the math be moved into inform_n (and warning_n) itself? No. I'm against

Re: [PATCH] Fix up one mishandled plural in diagnostics in gimple-ssa-sprintf.c

2017-12-19 Thread Jakub Jelinek
On Tue, Dec 19, 2017 at 11:19:44AM -0700, Martin Sebor wrote: > On 12/19/2017 10:58 AM, Jakub Jelinek wrote: > > On Tue, Dec 19, 2017 at 10:49:07AM -0700, Martin Sebor wrote: > > > Can the math be moved into inform_n (and warning_n) itself? > > > > No. I'm against having dozens of inform_n and

Re: [PATCH, PR83492] Fix selection of aarch64 big-endian shift parameters based on __AARCH64EB__

2017-12-19 Thread Jakub Jelinek
On Tue, Dec 19, 2017 at 07:01:44PM +0100, Michael Weiser wrote: > Hi Kyrill, > > On Tue, Dec 19, 2017 at 05:35:03PM +, Kyrill Tkachov wrote: > > > > below patch fixes PR 83492. > > I agree with your analysis of the bug and your patch will fix the > > problem when compiling with GCC. However,

Re: [PATCH] Fix up one mishandled plural in diagnostics in gimple-ssa-sprintf.c

2017-12-19 Thread Martin Sebor
On 12/19/2017 10:58 AM, Jakub Jelinek wrote: On Tue, Dec 19, 2017 at 10:49:07AM -0700, Martin Sebor wrote: Can the math be moved into inform_n (and warning_n) itself? No. I'm against having dozens of inform_n and warning_n etc. The question/suggestion is to change the existing inform_n and

Re: [SFN] Bootstrap broken

2017-12-19 Thread Alexandre Oliva
On Dec 15, 2017, Jakub Jelinek wrote: > Please don't revert the above 2 hunks. [...] We want to verify there are > no statements before labels. Why, sure! Thanks for catching this. -- Alexandre Oliva, freedom fighterhttp://FSFLA.org/~lxoliva/ You must be the change

Re: [PATCH, PR83492] Fix selection of aarch64 big-endian shift parameters based on __AARCH64EB__

2017-12-19 Thread Michael Weiser
Hi Kyrill, On Tue, Dec 19, 2017 at 05:35:03PM +, Kyrill Tkachov wrote: > > below patch fixes PR 83492. > I agree with your analysis of the bug and your patch will fix the > problem when compiling with GCC. However, I believe the more > appropriate macro to check here is __ARM_BIG_ENDIAN as

Re: [PATCH] Fix up one mishandled plural in diagnostics in gimple-ssa-sprintf.c

2017-12-19 Thread Jakub Jelinek
On Tue, Dec 19, 2017 at 07:02:38PM +0100, Richard Biener wrote: > On December 19, 2017 6:58:15 PM GMT+01:00, Jakub Jelinek > wrote: > >On Tue, Dec 19, 2017 at 10:49:07AM -0700, Martin Sebor wrote: > >> Can the math be moved into inform_n (and warning_n) itself? > > > >No. I'm

Re: [PATCH] Fix up one mishandled plural in diagnostics in gimple-ssa-sprintf.c

2017-12-19 Thread Richard Biener
On December 19, 2017 6:58:15 PM GMT+01:00, Jakub Jelinek wrote: >On Tue, Dec 19, 2017 at 10:49:07AM -0700, Martin Sebor wrote: >> Can the math be moved into inform_n (and warning_n) itself? > >No. I'm against having dozens of inform_n and warning_n etc. >overloads, it is

Re: [PATCH] Fix up one mishandled plural in diagnostics in gimple-ssa-sprintf.c

2017-12-19 Thread Jakub Jelinek
On Tue, Dec 19, 2017 at 10:49:07AM -0700, Martin Sebor wrote: > Can the math be moved into inform_n (and warning_n) itself? No. I'm against having dozens of inform_n and warning_n etc. overloads, it is already bad enough we have so many diagnostics entry points now that putting breakpoints on

Re: [PATCH] Small load strlen opt follow-up improvement (PR tree-optimization/83444)

2017-12-19 Thread Richard Biener
On December 19, 2017 6:23:55 PM GMT+01:00, Jakub Jelinek wrote: >Hi! > >As the testcase shows, we weren't handling the case where we have >_34 = MEM_REF[, 4B]; >- we would (if it exists) find the string length for b.a, but not >for the second field. Using get_addr_stridx

[PR83422] SFN: don't drop markers for skipping var-tracking

2017-12-19 Thread Alexandre Oliva
Although debug markers are more useful when bind stmts are placed among them, there is value in keeping them even when VTA limits are exceeded. Regstrapped on x86_64- and i686-linux-gnu. Preapproved by Jakub in the PR. Installed. for gcc/ChangeLog PR debug/83422 *

Re: [PATCH] Fix up one mishandled plural in diagnostics in gimple-ssa-sprintf.c

2017-12-19 Thread Martin Sebor
On 12/19/2017 10:20 AM, Jakub Jelinek wrote: Hi! This inform is correct in english, but wrong in many other languages. I think someone already mentioned this in another review but the computation below is far from intuitive: + ? (fmtres.range.likely % 100) + 100 +

Re: [compare-debug] use call loc for nop_endbr

2017-12-19 Thread Alexandre Oliva
On Dec 15, 2017, "Tsimbalist, Igor V" wrote: > Please give more info. I do not remember all details but PLT entries > were changes to have endbr instruction (if this is relevant to your question > :). Yeah, that was this was about. Good to know it's covered.

Re: [PATCH v3] Ability to remap file names in __FILE__, etc (PR other/70268)

2017-12-19 Thread Ximin Luo
Boris Kolpackov: > Thanks for the review. Third revision of the patch attached. > > [..] Here is a follow-up patch, meant to be applied on top of Boris' patch, to address some issues we found based on discussions and experiments at the Reproducible Builds project. Most of this patch is

Re: [PATCH, PR83492] Fix selection of aarch64 big-endian shift parameters based on __AARCH64EB__

2017-12-19 Thread Kyrill Tkachov
Hi Michael, [cc'ing the aarch64 maintainers] On 19/12/17 17:10, Michael Weiser wrote: Hello, below patch fixes PR 83492. Running the preprocessor on any source gives various errors on aarch64_be. These suggest that the proprocessor misidentifies token boundaries. This can be traced to an

Re: [PATCH] diagnose attribute conflicts on conversion operators (PR 83394)

2017-12-19 Thread Jason Merrill
On 12/18/2017 05:25 PM, Martin Sebor wrote: On 12/13/2017 02:38 PM, Jason Merrill wrote: On Wed, Dec 13, 2017 at 12:54 PM, Martin Sebor wrote: The attached update also fixes both instances of the ICE reported in bug 83322 and supersedes Jakub's patch for that bug

[PATCH] Small load strlen opt follow-up improvement (PR tree-optimization/83444)

2017-12-19 Thread Jakub Jelinek
Hi! As the testcase shows, we weren't handling the case where we have _34 = MEM_REF[, 4B]; - we would (if it exists) find the string length for b.a, but not for the second field. Using get_addr_stridx handles that well. The reason to use get_stridx on the MEM_REF first operand is to handle the

[PATCH] Fix up one mishandled plural in diagnostics in gimple-ssa-sprintf.c

2017-12-19 Thread Jakub Jelinek
Hi! This inform is correct in english, but wrong in many other languages. Bootstrapped/regtested on x86_64-linux and i686-linux, ok for trunk? 2017-12-19 Jakub Jelinek * gimple-ssa-sprintf.c (format_directive): Use inform_n instead of inform with hardcoded

[committed] Fix cswtch* tests on hppa64 (PR testsuite/83454)

2017-12-19 Thread Jakub Jelinek
Hi! The switchconv pass attempts to avoid creating rodata arrays that need runtime relocations, so for the tests we need non-pic. Tested on x86_64-linux and i686-linux, committed as obvious to trunk. 2017-12-19 Jakub Jelinek PR testsuite/83454 *

[og7] [PR other/79543] Fix GNU ld --version scanning to conform to the GNU Coding Standards

2017-12-19 Thread Thomas Schwinge
Hi! On Thu, 19 Oct 2017 18:35:09 +0200, I wrote: > As discussed in : [...] > [PR other/79543] Fix GNU ld --version scanning to conform to the GNU > Coding Standards I have not yet addressed Joseph's review comments (thanks!), but to allow Chung-Lin to make

[PATCH, PR83492] Fix selection of aarch64 big-endian shift parameters based on __AARCH64EB__

2017-12-19 Thread Michael Weiser
Hello, below patch fixes PR 83492. Running the preprocessor on any source gives various errors on aarch64_be. These suggest that the proprocessor misidentifies token boundaries. This can be traced to an optimized search_line_fast for aarch64 in libcpp/lex.c which has a special provision for

[PATCH][arm] PR target/82975: Guard against reg_renumber being NULL in arm.h

2017-12-19 Thread Kyrill Tkachov
Hi all, In this bug we ICE when checking REGNO_OK_FOR_INDEX_P on arm during pre-IRA scheduling. This is because REGNO_OK_FOR_INDEX_P ends up checking the reg_renumber array. Before IRA reg_renumber is NULL and thus we segfault. The fix is to guard the use of reg_renumber in the logic in

[PATCH, PR83423] Don't call targetm.calls.static_chain in non-static function

2017-12-19 Thread Tom de Vries
Hi, this patch fixes PR83423. The default_static_chain hook has implemented a sorry if both STATIC_CHAIN_INCOMING_REGNUM and STATIC_CHAIN_REGNUM are undefined: ... rtx default_static_chain (const_tree ARG_UNUSED (fndecl_or_type), bool incoming_p) { if (incoming_p) { #ifdef

Re: [PATCH 1/7]: SVE: Add CLOBBER_HIGH expression

2017-12-19 Thread Jeff Law
On 12/19/2017 03:12 AM, Alan Hayward wrote: > Ping ping. > I think there should be enough information in the first test to show that any > "set to self” > registers become live. Let me know if there’s anything I’ve missed. I think that both Richi and I would like you investigate fixing the df

Re: [PATCH] Add two test-cases for (PR middle-end/82404).

2017-12-19 Thread Jeff Law
On 12/19/2017 03:16 AM, Martin Liška wrote: > Hello. > > Following patch adds two test-cases for the PR. > > Ready for trunk? > Martin > > gcc/testsuite/ChangeLog: > > 2017-12-19 Martin Liska > > PR middle-end/82404 > * g++.dg/pr82404.C: New test. > *

[PATCH BACKPORT]Backport r254778 and test case in r244815 to GCC6

2017-12-19 Thread Bin Cheng
HI, This patch backports r254778 and test case in r244815 to GCC6. Bootstrap and test on x86_64. Is it OK? Thanks, bin 2017-12-18 Bin Cheng Backport from mainline 2017-11-15 Bin Cheng PR tree-optimization/82726 PR

Re: [GCC BACKPORT]Backport revision 254777 and 254778 to GCC 7 branch

2017-12-19 Thread Richard Biener
On Tue, Dec 19, 2017 at 4:15 PM, Bin Cheng wrote: > Hi, > This patch backports revision 254777 and 254778 to GCC 7 branch. > Bootstrap and test on x86_64. Is it OK? Ok. Richard. > Thanks, > bin > > 2017-12-18 Bin Cheng > > Backport from

[GCC BACKPORT]Backport revision 254777 and 254778 to GCC 7 branch

2017-12-19 Thread Bin Cheng
Hi, This patch backports revision 254777 and 254778 to GCC 7 branch. Bootstrap and test on x86_64. Is it OK? Thanks, bin 2017-12-18 Bin Cheng Backport from mainline 2017-11-15 Bin Cheng PR tree-optimization/82726 PR

[PATCH] [ARC][COMMITTED] Clean up build warnings.

2017-12-19 Thread Claudiu Zissulescu
Committed as obvious. //Claudiu 2017-12-19 Claudiu Zissulescu * config/arc/arc.c (legitimate_scaled_address_p): Clean fall-through warning. (arc_compute_frame_size): Remove unused variables. (arc_print_operand): Fix fprintif format.

Re: PATCH to fix ICE-on-invalid with -Wabi (PR c++/83490)

2017-12-19 Thread Jakub Jelinek
On Tue, Dec 19, 2017 at 03:50:33PM +0100, Marek Polacek wrote: > This is an ICE-on-invalid where the code in init_cumulative_args tries to > determine if it should warn about the empty classes ABI change, but is upset > when it encounters error_mark_node. Thus fixed. > > Bootstrapped/regtested

Re: [patch, doc] fix some overfull hboxes

2017-12-19 Thread Michel Morin
Hi, > Index: gcc/doc/invoke.texi [...] > +In C++14, @code{f} calls calls @code{f}, but in -> +In C++14, @code{f(g)} calls @code{f}, but in ? (Sorry for unthreading..., I haven't subscribed this ML yet and I couldn't set "In-Reply-To" header manually on Gmail web

PATCH to fix ICE-on-invalid with -Wabi (PR c++/83490)

2017-12-19 Thread Marek Polacek
This is an ICE-on-invalid where the code in init_cumulative_args tries to determine if it should warn about the empty classes ABI change, but is upset when it encounters error_mark_node. Thus fixed. Bootstrapped/regtested on x86_64-linux, ok for trunk? 2017-12-19 Marek Polacek

[PATCH] Fix PR83452

2017-12-19 Thread Richard Biener
The following enables a hpux specific workaround for the WEAK UNDEF symbols used by LTO debuginfo extraction. It makes sure to use a special name (gnu_lto_v1) for those. It also adjusts removed local symbols to not use UNDEFs (similar to solaris ld for globals hpux doesn't like undefs it cannot

Re: [PATCH] Replace Yoda conditions in gcc/

2017-12-19 Thread Richard Biener
On Tue, 19 Dec 2017, Jakub Jelinek wrote: > Hi! > > Segher's comment in one of the PRs > ('-- We don't use "yoda conditions" in GCC ("0 == x");') > I was looking at led me discover we have more Yoda conditions in the source > than healthy. > > This patch fixes the comparison order to the usual

[PATCH] Replace Yoda conditions in gcc/

2017-12-19 Thread Jakub Jelinek
Hi! Segher's comment in one of the PRs ('-- We don't use "yoda conditions" in GCC ("0 == x");') I was looking at led me discover we have more Yoda conditions in the source than healthy. This patch fixes the comparison order to the usual style. If some port maintainer used 1 <= x && x <= 24

[PATCH, rs6000] Don't optimize swaps when a swap has mixed use

2017-12-19 Thread Bill Schmidt
Hi, Carl Love is working on a patch to add missing flavors of the vec_xst_be intrinsic and test cases to cover all flavors. He ran into a latent bug in swap optimization that this patch addresses. Swap optimization operates on the principle that a computation can have swaps removed if all

Re: [i386] PR81842 [CET] -fcf-protection -mcet is incompatible with makecontext family functions

2017-12-19 Thread H.J. Lu
On Tue, Dec 19, 2017 at 4:41 AM, Andreas Schwab wrote: > On Dez 19 2017, "H.J. Lu" wrote: > >> This isn't sufficient. I did get >> >> ../sysdeps/unix/sysv/linux/x86/tst-ucontext-1a.c:25:10: error: call to >> \u2018getcontext\u2019 declared with attribute

Re: [patch][x86] -march=icelake

2017-12-19 Thread Jakub Jelinek
On Tue, Dec 19, 2017 at 12:34:03PM +, Koval, Julia wrote: > >> Maybe [] operator could be used instead of a dynamic handling here. > I had another solution in mind, with enums, which then addresses elements > using its index, please look the patch attached. You can also have a look at the

Re: [PATCH][GRAPHITE] Fix PR71351

2017-12-19 Thread Tom de Vries
On 12/19/2017 02:05 PM, Richard Biener wrote: On Tue, 19 Dec 2017, Tom de Vries wrote: On 09/21/2017 12:07 PM, Richard Biener wrote: -exit_edge = create_empty_if_region_on_edge (entry_edge, - unshare_expr (cond_expr)); This removes the fix

[PATCH] Add a tetst for PR middle-end/83471

2017-12-19 Thread H.J. Lu
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=83471 has been fixed on gcc-7-branch. I am checking this into master and gcc-7-branch. H.J. --- PR middle-end/83471 * g++.dg/torture/pr83471.C: New test. --- gcc/testsuite/g++.dg/torture/pr83471.C | 21 + 1 file

Re: Replying to an older patch ([PATCH] Fix failing test-case)

2017-12-19 Thread Martin Liška
On 12/19/2017 02:03 PM, Sudakshina Das wrote: > Hi Martin > > On 19/12/17 10:49, Martin Liška wrote: >> On 11/30/2017 12:03 PM, Sudakshina Das wrote: >>> https://gcc.gnu.org/ml/gcc-patches/2017-10/msg01157.html >>> >>> This patch fixed a test case switch-case-2.c. I am seeing switch-case-1.c >>>

Re: [patch][x86] -march=icelake

2017-12-19 Thread Richard Biener
On Tue, Dec 19, 2017 at 1:34 PM, Koval, Julia wrote: >>> Maybe [] operator could be used instead of a dynamic handling here. > I had another solution in mind, with enums, which then addresses elements > using its index, please look the patch attached. > > The natural

Re: [PATCH][GRAPHITE] Fix PR71351

2017-12-19 Thread Richard Biener
On Tue, 19 Dec 2017, Tom de Vries wrote: > On 09/21/2017 12:07 PM, Richard Biener wrote: > > -exit_edge = create_empty_if_region_on_edge (entry_edge, > > - unshare_expr (cond_expr)); > > This removes the fix for PR70045: > ... > diff --git

Re: Replying to an older patch ([PATCH] Fix failing test-case)

2017-12-19 Thread Sudakshina Das
Hi Martin On 19/12/17 10:49, Martin Liška wrote: On 11/30/2017 12:03 PM, Sudakshina Das wrote: https://gcc.gnu.org/ml/gcc-patches/2017-10/msg01157.html This patch fixed a test case switch-case-2.c. I am seeing switch-case-1.c failing on arm-none-linux-gnueabihf: FAIL:

Re: [PATCH PR81740]Enforce dependence check for outer loop vectorization

2017-12-19 Thread Richard Biener
On Tue, Dec 19, 2017 at 12:58 PM, Bin.Cheng wrote: > On Mon, Dec 18, 2017 at 2:35 PM, Michael Matz wrote: >> Hi, >> >> On Mon, 18 Dec 2017, Richard Biener wrote: >> >>> where *unroll is similar to *max_vf I think. dist_v[0] is the innermost >>> loop. >> >>

Re: [i386] PR81842 [CET] -fcf-protection -mcet is incompatible with makecontext family functions

2017-12-19 Thread Andreas Schwab
On Dez 19 2017, "H.J. Lu" wrote: > This isn't sufficient. I did get > > ../sysdeps/unix/sysv/linux/x86/tst-ucontext-1a.c:25:10: error: call to > \u2018getcontext\u2019 declared with attribute error: -mshstk > incompatible with ucontext.h APIs >return getcontext (ucp);

Re: [PATCH][GRAPHITE] Fix PR71351

2017-12-19 Thread Tom de Vries
On 09/21/2017 12:07 PM, Richard Biener wrote: -exit_edge = create_empty_if_region_on_edge (entry_edge, - unshare_expr (cond_expr)); This removes the fix for PR70045: ... diff --git a/gcc/graphite-isl-ast-to-gimple.c

RE: [patch][x86] -march=icelake

2017-12-19 Thread Koval, Julia
>> Maybe [] operator could be used instead of a dynamic handling here. I had another solution in mind, with enums, which then addresses elements using its index, please look the patch attached. >>> The natural GCC data structure is a sbitmap ... I'd rather not use >>> given we have a GCC

Re: [i386] PR81842 [CET] -fcf-protection -mcet is incompatible with makecontext family functions

2017-12-19 Thread H.J. Lu
On Tue, Dec 19, 2017 at 2:25 AM, Jakub Jelinek wrote: > On Tue, Dec 19, 2017 at 09:42:11AM +, Tsimbalist, Igor V wrote: >> > I don't see any discussion in the bugzilla issue to explain this. >> >> This option is needed to support two cases: >> >> 1. Compilation of ucontext

Re: [PATCH PR81740]Enforce dependence check for outer loop vectorization

2017-12-19 Thread Bin.Cheng
On Mon, Dec 18, 2017 at 2:35 PM, Michael Matz wrote: > Hi, > > On Mon, 18 Dec 2017, Richard Biener wrote: > >> where *unroll is similar to *max_vf I think. dist_v[0] is the innermost >> loop. > > [0] is always outermost loop. > >> The vectorizer does way more complicated things

C++ PATCH to add a test for c++/82231

2017-12-19 Thread Marek Polacek
This bug was fixed in r255430, but the test looks different enough so we better add it. Tested on x86_64-linux, ok for trunk? 2017-12-19 Marek Polacek PR c++/82231 * g++.dg/cpp1z/nontype-auto14.C: New test. --- gcc/testsuite/g++.dg/cpp1z/nontype-auto14.C

Re: [patch][x86] -march=icelake

2017-12-19 Thread Richard Biener
On Tue, Dec 19, 2017 at 9:29 AM, Uros Bizjak wrote: > On Mon, Dec 18, 2017 at 2:42 PM, Koval, Julia wrote: >> Hi, I tried to replace 2 flags variable with c++ bitset(in patch attached). >> What do you think? > > Hm, I'm not a c++ person, but I wonder

Re: [PATCH] Fix off by one error in loop-unroll.c (PR rtl-optimization/82675).

2017-12-19 Thread Richard Biener
On Tue, Dec 19, 2017 at 8:56 AM, Martin Liška wrote: > Hello. > > Following patch fixes off by one in loop unroller where wont_exit has > not enough elements that are later on asked for. > > Patch can bootstrap on ppc64le-redhat-linux and survives regression tests. > > Ready to be

  1   2   >