Re: Remove mode argument from gen_rtx_SET
DJ Delorie writes: >> ; This pattern is identical to the truncsipsi2 pattern except >> ; that it uses a SUBREG instead of a TRUNC. It is needed in >> ; order to prevent reload from converting (set:SI (SUBREG:PSI (SI))) >> ; into (SET:PSI (PSI)). >> >> I'm not sure what that's supposed to mean (what's an SI set of a PSI >> subreg?), but I suspect removing the mode would lose information, >> so I left it alone. > > MSP430 has 20-bit registers (PSImode-sized). One register can hold an > HI or PSI sized value, but if you have an SI value it's stored as two > HI registers. > > Thus, a PSImode value in a register is *not* just the 20 LSB of an > SImode value. Also, a PSImode subset of an SI value is stored > different than a PSImode value on its own. > > Thus, consider code like this: > > (set (reg:SI 1) > (subreg:PSI (reg:SI 2))) > > (set (reg:PSI 1) > (reg:PSI 2)) > > On most architectures, you'd say "these do the same thing" but on > MSP430 they don't. What I was confused about is that the first set isn't valid rtl. The SET_SRC and SET_DEST always have to have the same mode (or VOIDmode in the case of a CONST_INT, etc., where the mode is implicitly the same as the SET_DEST). So wouldn't it have to be: (set (reg:SI 1) (subreg:SI (reg:PSI 2))) or: (set (reg:PSI 1) (subreg:PSI (reg:SI 2))) ? Thanks, Richard
Re: [patch] defer C++ mangling aliases
> ... > I have tested it independently of everything else, > and would like permission to commit it to mainline. > ... This caused https://gcc.gnu.org/ml/gcc-bugs/2015-05/msg00748.html Dominique
Re: PATCHes to help with C++11 bootstrap
On May 9, 2015 6:30:49 AM GMT+02:00, Jason Merrill wrote: >One C++11 compatibility issue that turns up a lot in the GCC sources is > >that in C++98, > >#define BAR "bar" >const char *p = "foo"BAR; > >is well-formed, giving p the value "foobar". But in C++11 this is a >user-defined literal with the suffix BAR, which is ill-formed because >there is no BAR suffix defined. > >-Wc++11-compat didn't warn about this, which I'm fixing with the first >patch. > >The second patch fixes all the occurrences in GCC. > >The third patch fixes the warning to say "-Wc++11-compat" rather than >"-Wc++0x-compat". > >The fourth patch fixes a few G++ tests that were failing with the >compiler defaulting to C++11. > >Tested x86_64-cp-linux-gnu, applying to trunk. Hmm, I wonder if we want to bootstrap with explicit -std=gnu04, our host compiler requirement. Otherwise we'll silently sneak in C++11 features when that becomes the default? Richard.
Re: [patch 1/10] debug-early merge: Ada front-end
> My question exactly. Perhaps that was my confusion. Why is this using > ->global_decl? Because that's equivalent to ->type_decl (iter, false) in DWARF. > For example, the C front-end uses rest_of_type_compilation (see > finish_struct() in c/c-decl.c) which calls ->type_decl(), or it calls > ->type_decl() from record_builtin_type(). All right, let's follow the gospel then. :-) Patch attached and installed. > I can put the code back, but calling ->type_decl()? Assuming you folks > (Ada) don't want to use rest_of_type_compilation(). No, we don't use it for local types either (it's a no-op in DWARF in this case and we don't support STABS anymore because of the need to manually delay the generation of debug info for local types, which is a real pain in Ada). * gcc-interface/utils.c (gnat_write_global_declarations): Use type_decl method instead of global_decl for TYPE_DECLs. -- Eric BotcazouIndex: gcc-interface/utils.c === --- gcc-interface/utils.c (revision 222673) +++ gcc-interface/utils.c (working copy) @@ -5211,7 +5211,7 @@ gnat_write_global_declarations (void) finalized in the right context. */ FOR_EACH_VEC_SAFE_ELT (global_decls, i, iter) if (TREE_CODE (iter) == TYPE_DECL && !DECL_IGNORED_P (iter)) - debug_hooks->global_decl (iter); + debug_hooks->type_decl (iter, false); /* Proceed to optimize and emit assembly. */ symtab->finalize_compilation_unit ();
Re: [patch 1/10] debug-early merge: Ada front-end
> Ah, I see what you mean. I'll address this and repost. Thanks. Did the first version pass the GDB testsuite? I thought we had a testcase exercising the pointer to Taft amendment type case in it. -- Eric Botcazou
Re: [patch, Fortran] Fix PR 66041
Hello,
Le 09/05/2015 00:11, Thomas Koenig a écrit :
> Index: frontend-passes.c
> ===
> --- frontend-passes.c (Revision 222864)
> +++ frontend-passes.c (Arbeitskopie)
> @@ -2611,14 +2611,40 @@ scalarized_expr (gfc_expr *e_in, gfc_expr **index,
> {
> /* Look at full individual sections, like a(:). The
> first index
>is the lbound of a full ref. */
> -
> + int j;
> gfc_array_ref *ar;
> + gfc_expr *lbound_e;
>
> - ar = gfc_find_array_ref (e_in);
> + lbound_e = gfc_copy_expr (e_in);
> + ar = gfc_find_array_ref (lbound_e);
> +
> ar->type = AR_FULL;
> + for (j = 0; j < ar->dimen; j++)
> + {
> + gfc_free_expr (ar->start[j]);
> + ar->start[j] = NULL;
> + gfc_free_expr (ar->end[j]);
> + ar->end[j] = NULL;
> + gfc_free_expr (ar->stride[j]);
> + ar->stride[j] = NULL;
> + }
> +
You also need to remove/free the trailing subreferences.
> + /* We have to get rid of the shape, if thre is one. Do
> + so by freeing it and calling gfc_resolve to rebuild it,
> + if necessary. */
> +
> + if (lbound_e->shape)
> + gfc_free_shape (&(lbound_e->shape), lbound_e->rank);
> +
> + lbound_e->rank = ar->dimen;
ar->dimen is not what you think it is.
It is 3 for array(1, 1, :), while the rank is 1.
gfc_resolve_expr should set the rank for you, so just remove this line.
> +
> + gfc_resolve_expr (lbound_e);
> + lbound = get_array_inq_function (GFC_ISYM_LBOUND,
> +lbound_e, i + 1);
free lbound_e?
> }
> - lbound = get_array_inq_function (GFC_ISYM_LBOUND, e_in,
> -i_index + 1);
> + else
> + lbound = get_array_inq_function (GFC_ISYM_LBOUND, e_in,
> + i_index + 1);
You can't reuse e_in if it has subreferences.
One suggestion: you may want to move all the above to a function
extracting the full array.
> }
>
> ar->dimen_type[i] = DIMEN_ELEMENT;
> @@ -2639,6 +2665,8 @@ scalarized_expr (gfc_expr *e_in, gfc_expr **index,
> i_index ++;
> }
> }
> + gfc_free_expr (e_in);
> +
This side effect is asking for trouble.
Instead of this, remove the copies made in the callers.
This is independant from the rest, so it can be made later as a follow-up.
>return e;
> }
>
>
I attach a variant of your inline_matmul_8.f90 that is not working yet
because of subreferences.
Mikael
--- inline_matmul_8.f90.old 2015-05-09 13:31:28.420790646 +0200
+++ inline_matmul_8.f90 2015-05-09 13:42:50.741799982 +0200
@@ -3,15 +3,22 @@
! PR 66041 - this used to ICE with an incomplete fix for the PR.
program main
implicit none
- real, dimension(1,-2:0) :: a1
+ type :: t
+real :: c
+ end type t
+ type(t), dimension(1,-2:0) :: a1
real, dimension(3,2) :: b1
real, dimension(2) :: c1
+ real, dimension(1,2) :: c2
- data a1 /17., -23., 29./
+ data a1%c /17., -23., 29./
data b1 / 2., -3., 5., -7., 11., -13./
- c1 = matmul(a1(1,:), b1)
+ c1 = matmul(a1(1,:)%c, b1)
if (any (c1-[248., -749.] /= 0.)) call abort
+
+ c2 = matmul(a1%c, b1)
+ if (any (c2-reshape([248., -749.],shape(c2)) /= 0.)) call abort
end program main
! { dg-final { scan-tree-dump-times "_gfortran_matmul" 0 "original" } }
Re: conditional lim
Evgeniya Maenkova writes: > > So, in my opinion it’s ok to generate additional phi node for debug > case. But I’m not a compiler expert and maybe there is some > requirement that debug and non-debug versions should differ only by > debug statements, I don’t know. gcc has such a requirement. Otherwise you cannot debug code that was originally not compiled with debugging. -Andi -- a...@linux.intel.com -- Speaking for myself only
Re: [Patch, fortran] PR65792 - allocation of scalar elemental function with structure constructor fails
Le 01/05/2015 20:25, Paul Richard Thomas a écrit :
> Dear All,
>
> By the time I went to commit, something had changed and the patch
> caused a regression. I presume that the version that I had of Andre's
> patch was not the same as the one committed. I'll cast an eye over it
> this weekend and see if I can understand what gives.
>
Hello Paul,
to get things moving again, I propose the attached fix to your patch.
Tested on alloc_comp_deep_copy_1 only for now.
Mikael
diff --git a/gcc/fortran/trans-expr.c b/gcc/fortran/trans-expr.c
index 6cf5fe2..532f4b7 100644
--- a/gcc/fortran/trans-expr.c
+++ b/gcc/fortran/trans-expr.c
@@ -7078,7 +7078,8 @@ gfc_trans_subcomponent_assign (tree dest, gfc_component * cm, gfc_expr * expr,
}
gfc_add_modify (&block, dest,
fold_convert (TREE_TYPE (dest), se.expr));
- if (cm->ts.u.derived->attr.alloc_comp)
+ if (cm->ts.u.derived->attr.alloc_comp
+ && expr->expr_type != EXPR_NULL)
{
tmp = gfc_copy_alloc_comp (cm->ts.u.derived, se.expr,
dest, expr->rank);
Re: [patch] defer C++ mangling aliases
On 05/09/2015 02:53 AM, Dominique Dhumieres wrote:
...
I have tested it independently of everything else,
and would like permission to commit it to mainline.
...
This caused https://gcc.gnu.org/ml/gcc-bugs/2015-05/msg00748.html
Dominique
Fixed with the following patch.
Committed as obvious.
commit 3e820aa6675e608371ccf4f2ae691baa15a4be7f
Author: Aldy Hernandez
Date: Sat May 9 06:47:33 2015 -0700
PR bootstrap/66085
* decl2.c (note_mangling_alias): Declare arguments as unused.
diff --git a/gcc/cp/decl2.c b/gcc/cp/decl2.c
index 6101301..27bacab 100644
--- a/gcc/cp/decl2.c
+++ b/gcc/cp/decl2.c
@@ -4327,7 +4327,7 @@ generate_mangling_alias (tree decl, tree id2)
implementation. */
void
-note_mangling_alias (tree decl, tree id2)
+note_mangling_alias (tree decl ATTRIBUTE_UNUSED, tree id2 ATTRIBUTE_UNUSED)
{
#ifdef ASM_OUTPUT_DEF
if (at_eof)
Re: PATCH] PR target/65612: Multiversioning doesn't work with DSO nor PIE
* H.J. Lu [2015-04-17 05:36:30 -0700]: > On Fri, Apr 17, 2015 at 4:59 AM, Jakub Jelinek wrote: > > On Fri, Apr 17, 2015 at 04:48:48AM -0700, H.J. Lu wrote: > >> > I don't like it. Nonshared libgcc is libgcc.a, period. No sense in > >> > creating yet another library for that. > >> > So, IMHO beyond making the __cpu* entrypoints compat symbols only (@ > >> > instead > >> > of @@ symbol versions) the right fix is simply tweak init_gcc_spec, so > >> > that > >> > static_name is always linked in, in the switch combinations that it isn't > >> > right now of course after shared_name rather than before that. > >> > I thought we've fixed that years ago... > >> > I think the patch committed for this is suboptimal. (it breaks with musl libc on x86 if libgcc_s is linked into a binary) the patch: http://gcc.gnu.org/ml/gcc-patches/2015-04/msg00878.html original thread: http://gcc.gnu.org/ml/gcc-patches/2015-03/msg01520.html The symbol versioning hack for __cpu_model and __cpu_indicator_init makes them invisible to the musl dynamic linker so their relocation fails with 'symbol not found' error. (affects anything linked with -lgcc_s) I don't yet understand all the constraints, what I gathered is * x86 multi-versioning was broken with libgcc_s because ifunc resolution tried to use a libgcc_s function before it was relocated. (mv dispatch uses ifunc before relocation is finished, that calls __builtin_cpu_init that turns into a call to __cpu_indicator_init that is a constructor in libgcc_s.so to initialize __cpu_model). * old binaries that happened to work still have to work with new libgcc_s.so so __cpu_* must have a definition there. * new binaries must get their own definition of __cpu_* (ie statically linked from libgcc.a) to avoid the ifunc issue. I would rename the __cpu_* symbols and put those only in libgcc.a while keeping the old __cpu_* symbols still available in libgcc_s.so: old binaries work with new gcc, new binaries get the right definitions (so they work on old systems and new systems too, and yes -lgcc should be passed to c++ dso linking). Did i miss something? (This is not in the gcc-5 branch so I assume It is OK to work on a less intrusive fix.) PS.: It would be nice if there was a specification for the ifunc interface contract before any ifunc/multiversioned interface gets into the compiler runtime (see libatomic.so) to avoid similar bugs and incompatibilities between the dynamic linker/loader and compiler, the dynamic linker cannot guarantee arbitrary ifunc calls to work. > >> > >> We never pass -lgcc to linker when building C++ DSO: > >> > >> /usr/libexec/gcc/x86_64-redhat-linux/4.9.2/collect2 -plugin > >> /usr/libexec/gcc/x86_64-redhat-linux/4.9.2/liblto_plugin.so > >> -plugin-opt=/usr/libexec/gcc/x86_64-redhat-linux/4.9.2/lto-wrapper > >> -plugin-opt=-fresolution=/tmp/ccZC7iqy.res > >> -plugin-opt=-pass-through=-lgcc_s -plugin-opt=-pass-through=-lc > >> -plugin-opt=-pass-through=-lgcc_s --build-id --no-add-needed > >> --eh-frame-hdr --hash-style=gnu -m elf_x86_64 -shared > >> /usr/lib/gcc/x86_64-redhat-linux/4.9.2/../../../../lib64/crti.o > >> /usr/lib/gcc/x86_64-redhat-linux/4.9.2/crtbeginS.o > >> -L/usr/lib/gcc/x86_64-redhat-linux/4.9.2 > >> -L/usr/lib/gcc/x86_64-redhat-linux/4.9.2/../../../../lib64 > >> -L/lib/../lib64 -L/usr/lib/../lib64 > >> -L/usr/lib/gcc/x86_64-redhat-linux/4.9.2/../../.. x.o -lstdc++ -lm > >> -lgcc_s -lc -lgcc_s /usr/lib/gcc/x86_64-redhat-linux/4.9.2/crtendS.o > >> /usr/lib/gcc/x86_64-redhat-linux/4.9.2/../../../../lib64/crtn.o > >> [hjl@gnu-32 tmp]$ > >> > >> That is why libgcc_nonshared.a is needed. > > > > See what I wrote. I think it is a bug that we don't do that, in your case > > we should pass -lgcc_s -lgcc -lc -lgcc_s -lgcc. > > Or, if you don't want to change that, as the multi-versioning change is > > i386/x86_64 only change, just ensure that those targets have > > t-slibgcc-libgcc in libgcc/config.host and thus behave like most other linux > > targets where -lgcc is linked in always after -lgcc_s. > > > > Jakub > > This patch works for me. OK for trunk? > > gcc/testsuite/ > > PR target/65612 > * g++.dg/ext/mv18.C: New test. > * g++.dg/ext/mv19.C: Likewise. > * g++.dg/ext/mv20.C: Likewise. > * g++.dg/ext/mv21.C: Likewise. > * g++.dg/ext/mv22.C: Likewise. > * g++.dg/ext/mv23.C: Likewise. > > libgcc/ > > PR target/65612 > * config.host (tmake_file): Add t-slibgcc-libgcc for Linux/x86. > * config/i386/cpuinfo.c (__cpu_model): Initialize. > (__cpu_indicator_init@GCC_4.8.0): New. > (__cpu_model@GCC_4.8.0): Likewise. > * config/i386/t-linux (HOST_LIBGCC2_CFLAGS): Add > -DUSE_ELF_SYMVER. > > Thanks. > > -- > H.J.
Re: [RFC][PATCH][X86_64] Eliminate PLT stubs for specified external functions via -fno-plt=
On Mon, May 4, 2015 at 7:45 AM, Michael Matz wrote: > Hi, > > On Thu, 30 Apr 2015, Sriraman Tallam wrote: > >> We noticed that one of our benchmarks sped-up by ~1% when we eliminated >> PLT stubs for some of the hot external library functions like memcmp, >> pow. The win was from better icache and itlb performance. The main >> reason was that the PLT stubs had no spatial locality with the >> call-sites. I have started looking at ways to tell the compiler to >> eliminate PLT stubs (in-effect inline them) for specified external >> functions, for x86_64. I have a proposal and a patch and I would like to >> hear what you think. >> >> This comes with caveats. This cannot be generally done for all >> functions marked extern as it is impossible for the compiler to say if a >> function is "truly extern" (defined in a shared library). If a function >> is not truly extern(ends up defined in the final executable), then >> calling it indirectly is a performance penalty as it could have been a >> direct call. > > This can be fixed by Alans idea. > >> Further, the newly created GOT entries are fixed up at >> start-up and do not get lazily bound. > > And this can be fixed by some enhancements in the linker and dynamic > linker. The idea is to still generate a PLT stub and make its GOT entry > point to it initially (like a normal got.plt slot). Then the first > indirect call will use the address of PLT entry (starting lazy resolution) > and update the GOT slot with the real address, so further indirect calls > will directly go to the function. > > This requires a new asm marker (and hence new reloc) as normally if > there's a GOT slot it's filled by the real symbols address, unlike if > there's only a got.plt slot. E.g. a > > call *foo@GOTPLT(%rip) > > would generate a GOT slot (and fill its address into above call insn), but > generate a JUMP_SLOT reloc in the final executable, not a GLOB_DAT one. > I added the "relax" prefix support to x86 assembler on users/hjl/relax branch at https://sourceware.org/git/?p=binutils-gdb.git;a=summary [hjl@gnu-tools-1 relax-3]$ cat r.S .text relax jmp foo relax call foo relax jmp foo@plt relax call foo@plt [hjl@gnu-tools-1 relax-3]$ ./as -o r.o r.S [hjl@gnu-tools-1 relax-3]$ ./objdump -drw r.o r.o: file format elf64-x86-64 Disassembly of section .text: <.text>: 0: 66 e9 00 00 00 00 data16 jmpq 0x6 2: R_X86_64_RELAX_PC32 foo-0x4 6: 66 e8 00 00 00 00 data16 callq 0xc 8: R_X86_64_RELAX_PC32 foo-0x4 c: 66 e9 00 00 00 00 data16 jmpq 0x12 e: R_X86_64_RELAX_PLT32foo-0x4 12: 66 e8 00 00 00 00 data16 callq 0x18 14: R_X86_64_RELAX_PLT32foo-0x4 [hjl@gnu-tools-1 relax-3]$ Right now, the relax relocations are treated as PC32/PLT32 relocations. I am working on linker support. -- H.J.
[RFC, PATCH] nonzero attribute, static array parameter
Hi,
here is a tentative patch to implement a new attribute nonzero,
which is similar to nonnull, but is not a function attribute
but a type attribute.
One reason is that nonnull is awkward to use. For this reason,
clang allows the use of nonnull in function parameters, but this
is incompatible with old and current use of this attribute in gcc
(though in a rather obscure use case).
See: https://gcc.gnu.org/ml/gcc/2015-05/msg00077.html
The other reason is that a nonzero type attribute is conceptually
much simpler and at the same time more general than the existing
nonnull and nonnull_return function attributes (and could replace
both), e.g. we can define non-zero types and diagnose all stores
of known constant 0 to variables of this type, use this for
computing better value ranges, etc.
This patch implements the attribute and adds a warning if a
zero constant is passed as argument with nonzero type attribute.
Also infer_nonnull_range in gcc/gimple.c is extended to make use
of the attribute (which in turn is used by ubsan).
In addition, in C the attribute is automatically added to
pointers declared as array parameters with static, e.g:
void foo(int a[static 5]);
Martin
(tested on x86_64-unknown-linux-gnu)
diff --git a/gcc/ChangeLog b/gcc/ChangeLog
index 77d9352..2f79344 100644
--- a/gcc/ChangeLog
+++ b/gcc/ChangeLog
@@ -1,3 +1,8 @@
+2015-05-09 Martin Uecker
+
+ * gcc/gimple.c (infer_nonnull_range): Process nonzero attribute.
+ * doc/invoki.texi
+
2015-05-08 Jim Wilson
* doc/install.texi (--enable-languages): Add missing jit and lto info.
diff --git a/gcc/c-family/ChangeLog b/gcc/c-family/ChangeLog
index 57f83c9..809ecf1 100644
--- a/gcc/c-family/ChangeLog
+++ b/gcc/c-family/ChangeLog
@@ -1,3 +1,9 @@
+2015-05-09 Martin Uecker
+
+ * c-common.c (handle_nonzero_attribute): New.
+ (check_function_nonzero): New.
+ (check_nonzero_arg): New.
+
2015-05-08 Marek Polacek
PR c/64918
diff --git a/gcc/c-family/c-common.c b/gcc/c-family/c-common.c
index 378f237..46fe749 100644
--- a/gcc/c-family/c-common.c
+++ b/gcc/c-family/c-common.c
@@ -380,6 +380,7 @@ static tree handle_deprecated_attribute (tree *, tree, tree, int,
static tree handle_vector_size_attribute (tree *, tree, tree, int,
bool *);
static tree handle_nonnull_attribute (tree *, tree, tree, int, bool *);
+static tree handle_nonzero_attribute (tree *, tree, tree, int, bool *);
static tree handle_nothrow_attribute (tree *, tree, tree, int, bool *);
static tree handle_cleanup_attribute (tree *, tree, tree, int, bool *);
static tree handle_warn_unused_result_attribute (tree *, tree, tree, int,
@@ -407,6 +408,7 @@ static tree handle_bnd_instrument (tree *, tree, tree, int, bool *);
static void check_function_nonnull (tree, int, tree *);
static void check_nonnull_arg (void *, tree, unsigned HOST_WIDE_INT);
+static void check_nonzero_arg (void *, tree, unsigned HOST_WIDE_INT);
static bool nonnull_check_p (tree, unsigned HOST_WIDE_INT);
static bool get_nonnull_operand (tree, unsigned HOST_WIDE_INT *);
static int resort_field_decl_cmp (const void *, const void *);
@@ -751,6 +753,8 @@ const struct attribute_spec c_common_attribute_table[] =
handle_tls_model_attribute, false },
{ "nonnull",0, -1, false, true, true,
handle_nonnull_attribute, false },
+ { "nonzero",0, -1, false, true, false,
+ handle_nonzero_attribute, false },
{ "nothrow",0, 0, true, false, false,
handle_nothrow_attribute, false },
{ "may_alias", 0, 0, false, true, false, NULL, false },
@@ -8940,6 +8944,26 @@ handle_vector_size_attribute (tree *node, tree name, tree args,
return NULL_TREE;
}
+
+/* Handle the "nonzero" attribute. */
+static tree
+handle_nonzero_attribute (tree *node, tree ARG_UNUSED (name),
+ tree ARG_UNUSED (args), int ARG_UNUSED (flags),
+ bool *no_add_attrs)
+{
+ tree n = *node;
+
+ if ((TREE_CODE (n) != POINTER_TYPE)
+ && (TREE_CODE (n) != INTEGER_TYPE))
+{
+ warning (OPT_Wattributes, "%qE attribute ignored", name);
+ *no_add_attrs = true;
+}
+
+ return NULL_TREE;
+}
+
+
/* Handle the "nonnull" attribute. */
static tree
handle_nonnull_attribute (tree *node, tree ARG_UNUSED (name),
@@ -9016,6 +9040,34 @@ handle_nonnull_attribute (tree *node, tree ARG_UNUSED (name),
return NULL_TREE;
}
+
+static void
+check_function_nonzero (const_tree fntype, int nargs, tree *argarray)
+{
+ if (TREE_CODE (fntype) != FUNCTION_TYPE) // C++ lambdas
+return;
+
+ function_args_iterator iter;
+ tree type;
+ int i = 0;
+
+ FOREACH_FUNCTION_ARGS (fntype, type, iter)
+{
+ if (i == nargs)
+ break;
+
+ tree a = lookup_attribute ("nonzero", TYPE_ATTRIBUTES (type));
+
+ if (a != NULL_TREE)
+check_function_arguments_recurse (check_nonzero_arg, NULL,
+ argarray[i], i + 1);
+
+ i++;
+ }
+}
+
+
+
/* Check the argument list of a function call
Re: PATCH] PR target/65612: Multiversioning doesn't work with DSO nor PIE
On Sat, May 9, 2015 at 7:31 AM, Szabolcs Nagy wrote: > * H.J. Lu [2015-04-17 05:36:30 -0700]: >> On Fri, Apr 17, 2015 at 4:59 AM, Jakub Jelinek wrote: >> > On Fri, Apr 17, 2015 at 04:48:48AM -0700, H.J. Lu wrote: >> >> > I don't like it. Nonshared libgcc is libgcc.a, period. No sense in >> >> > creating yet another library for that. >> >> > So, IMHO beyond making the __cpu* entrypoints compat symbols only (@ >> >> > instead >> >> > of @@ symbol versions) the right fix is simply tweak init_gcc_spec, so >> >> > that >> >> > static_name is always linked in, in the switch combinations that it >> >> > isn't >> >> > right now of course after shared_name rather than before that. >> >> > I thought we've fixed that years ago... >> >> > > > I think the patch committed for this is suboptimal. > (it breaks with musl libc on x86 if libgcc_s is linked into a binary) > > the patch: > http://gcc.gnu.org/ml/gcc-patches/2015-04/msg00878.html > original thread: > http://gcc.gnu.org/ml/gcc-patches/2015-03/msg01520.html > > The symbol versioning hack for __cpu_model and __cpu_indicator_init > makes them invisible to the musl dynamic linker so their relocation > fails with 'symbol not found' error. > (affects anything linked with -lgcc_s) There are 4: 2b70 806 FUNCGLOBAL DEFAULT 12 __cpu_indicator_init@GCC_4.8.0 38: 002153d016 OBJECT GLOBAL DEFAULT 25 __cpu_model@GCC_4.8.0 and 0215000 00040001 R_X86_64_64 2b70 __cpu_indicator_init@GCC_4.8.0 + 0 00215220 00260006 R_X86_64_GLOB_DAT 002153d0 __cpu_model@GCC_4.8.0 + 0 in libgcc_s.so.1. Musl ld.so must be fixed to handle it. -- H.J.
[PATCH] D demangle: Correctly decode white or non-printable characters
D templates can have string literals encoded inside them, which can
also include tabs, newlines, and other whitespace characters. For
example:
return getHost!q{
auto he = gethostbyname(toStringz(param));
}(name);
In this case, rather than decoding and writing out every character
directly, whitespace or other non-printable characters should be
represented as escape sequences.
---
libiberty/ChangeLog:
2015-05-09 Iain Buclaw
* d-demangle.c (dlang_parse_string): Represent embedded whitespace or
non-printable characters as hex or escape sequences.
* testsuite/d-demangle-expected: Add test for templates with tabs and
newlines embedded into the signature.
---
diff --git a/libiberty/d-demangle.c b/libiberty/d-demangle.c
index bb481c0..fa01767 100644
--- a/libiberty/d-demangle.c
+++ b/libiberty/d-demangle.c
@@ -931,7 +931,38 @@ dlang_parse_string (string *decl, const char *mangled)
char a = ascii2hex (mangled[0]);
char b = ascii2hex (mangled[1]);
char val = (a << 4) | b;
- string_appendn (decl, &val, 1);
+
+ /* Sanitize white and non-printable characters. */
+ switch (val)
+ {
+ case ' ':
+ string_append (decl, " ");
+ break;
+ case '\t':
+ string_append (decl, "\\t");
+ break;
+ case '\n':
+ string_append (decl, "\\n");
+ break;
+ case '\r':
+ string_append (decl, "\\r");
+ break;
+ case '\f':
+ string_append (decl, "\\f");
+ break;
+ case '\v':
+ string_append (decl, "\\v");
+ break;
+
+ default:
+ if (ISPRINT (val))
+ string_appendn (decl, &val, 1);
+ else
+ {
+ string_append (decl, "\\x");
+ string_appendn (decl, mangled, 2);
+ }
+ }
}
else
return NULL;
diff --git a/libiberty/testsuite/d-demangle-expected b/libiberty/testsuite/d-demangle-expected
index 2aeacb8..b023f6d 100644
--- a/libiberty/testsuite/d-demangle-expected
+++ b/libiberty/testsuite/d-demangle-expected
@@ -934,3 +934,7 @@ serenity.persister.Sqlite.SqlitePersister!(serenity.persister.Sqlite.__unittest6
--format=dlang
_D4test4mainFZv5localMFZi
test.main().local()
+#
+--format=dlang
+_D3std6socket12InternetHost221__T13getHostNoSyncVAyaa96_0a09202020206175746f2078203d2068746f6e6c28706172616d293b0a09202020206175746f206865203d20676574686f73746279616464722826782c20342c206361737428696e74294164647265737346616d696c792e494e4554293b0a09TkZ13getHostNoSyncMFkZb
+std.socket.InternetHost.getHostNoSync!("\n\tauto x = htonl(param);\n\tauto he = gethostbyaddr(&x, 4, cast(int)AddressFamily.INET);\n\t", uint).getHostNoSync(uint)
[PATCH] D demangle: Include type modifiers in demangled function symbols
Like C++ const and volatile, in D mangled symbols can exist modifiers
that represent the const, immutable, inout and shared-ness of the
'this' parameter.
This information should be written out in the demangled symbol to show
that each variant has a unique identity.
---
libiberty/ChangeLog:
2015-05-09 Iain Buclaw
* d-demangle.c (dlang_type_modifiers): New function.
(dlang_type_modifier_p): New function.
(dlang_call_convention_p): Ignore any kind of type modifier.
(dlang_parse_symbol): Emit the type modifier after the symbol.
* testsuite/d-demangle-expected: Add coverage tests for all valid
usages of functions symbols with type modifiers.
---
diff --git a/libiberty/d-demangle.c b/libiberty/d-demangle.c
index bb481c0..9e6c46f 100644
--- a/libiberty/d-demangle.c
+++ b/libiberty/d-demangle.c
@@ -692,6 +692,44 @@ dlang_identifier (string *decl, const char *mangled)
return mangled;
}
+/* Extract the type modifiers from MANGLED and append them to DECL.
+ Returns the remaining signature on success or NULL on failure. */
+static const char *
+dlang_type_modifiers (string *decl, const char *mangled)
+{
+ if (mangled == NULL || *mangled == '\0')
+return mangled;
+
+ switch (*mangled)
+{
+case 'x': /* const */
+ mangled++;
+ string_append (decl, " const");
+ return mangled;
+case 'y': /* immutable */
+ mangled++;
+ string_append (decl, " immutable");
+ return mangled;
+case 'O': /* shared */
+ mangled++;
+ string_append (decl, " shared");
+ return dlang_type_modifiers (decl, mangled);
+case 'N':
+ mangled++;
+ if (*mangled == 'g') /* wild */
+ {
+ mangled++;
+ string_append (decl, " inout");
+ return dlang_type_modifiers (decl, mangled);
+ }
+ else
+ return NULL;
+
+default:
+ return mangled;
+}
+}
+
/* Extract the integer value from MANGLED and append it to DECL,
where TYPE is the type it should be represented as.
Return the remaining string on success or NULL on failure. */
@@ -1104,6 +1142,41 @@ dlang_value (string *decl, const char *mangled, const char *name, char type)
return mangled;
}
+/* Extract the type modifiers from MANGLED and return the string
+ length that it consumes in MANGLED on success or 0 on failure. */
+static int
+dlang_type_modifier_p (const char *mangled)
+{
+ int i;
+
+ switch (*mangled)
+{
+case 'x': case 'y':
+ return 1;
+
+case 'O':
+ mangled++;
+ i = dlang_type_modifier_p (mangled);
+ return i + 1;
+
+case 'N':
+ mangled++;
+ if (*mangled == 'g')
+ {
+ mangled++;
+ i = dlang_type_modifier_p (mangled);
+ return i + 2;
+ }
+
+default:
+ break;
+}
+
+ return 0;
+}
+
+/* Extract the function calling convention from MANGLED and
+ return 1 on success or 0 on failure. */
static int
dlang_call_convention_p (const char *mangled)
{
@@ -1116,9 +1189,9 @@ dlang_call_convention_p (const char *mangled)
return 1;
case 'M': /* Prefix for functions needing 'this' */
- i = 1;
- if (mangled[i] == 'x')
- i++;
+ mangled++;
+ /* Skip over any type modifiers. */
+ i = dlang_type_modifier_p (mangled);
switch (mangled[i])
{
@@ -1147,11 +1220,16 @@ dlang_parse_symbol (string *decl, const char *mangled)
if (mangled && dlang_call_convention_p (mangled))
{
+ string mods;
int saved;
/* Skip over 'this' parameter. */
if (*mangled == 'M')
- mangled += (mangled[1] == 'x') ? 2 : 1;
+ mangled++;
+
+ /* Save the type modifiers for appending at the end. */
+ string_init (&mods);
+ mangled = dlang_type_modifiers (&mods, mangled);
/* Skip over calling convention and attributes in qualified name. */
saved = string_length (decl);
@@ -1170,6 +1248,10 @@ dlang_parse_symbol (string *decl, const char *mangled)
mangled = dlang_type (decl, mangled);
string_setlength (decl, saved);
}
+
+ /* Add any const/immutable/shared modifier. */
+ string_appendn (decl, mods.b, string_length (&mods));
+ string_delete (&mods);
}
}
while (mangled && ISDIGIT (*mangled));
diff --git a/libiberty/testsuite/d-demangle-expected b/libiberty/testsuite/d-demangle-expected
index 2aeacb8..358372a 100644
--- a/libiberty/testsuite/d-demangle-expected
+++ b/libiberty/testsuite/d-demangle-expected
@@ -753,6 +753,38 @@ demangle.test!(demangle.S(1, 2))
_D8demangle35__T4testVS8demangle1SS2i1a3_616263Zv
demangle.test!(demangle.S(1, "abc"))
#
+--format=dlang
+_D8demangle4testMxFZv
+demangle.test() const
+#
+--format=dlang
+_D8demangle4testMyFZv
+demangle.test() immutable
+#
+--format=dlang
+_D8demangle4testMNgFZv
+demangle.test() inout
+#
+--format=dlang
+_D8demangle4testMNgxFZv
+demangle.test() inout const
+#
+--format=dlang
+_D8demangle4testMOFZv
+demangle.test() shared
+#
+--format=dlang
+_D8demangle4testMOxFZv
+demangle.test() shared const
+#
+--format=dlang
+_D8demangle4testMONgFZv
+demangle.
Re: PATCHes to help with C++11 bootstrap
On 2015.05.08 at 23:30 -0500, Jason Merrill wrote:
> One C++11 compatibility issue that turns up a lot in the GCC sources is
> that in C++98,
>
> #define BAR "bar"
> const char *p = "foo"BAR;
>
> is well-formed, giving p the value "foobar". But in C++11 this is a
> user-defined literal with the suffix BAR, which is ill-formed because
> there is no BAR suffix defined.
>
> -Wc++11-compat didn't warn about this, which I'm fixing with the first
> patch.
>
> The second patch fixes all the occurrences in GCC.
>
> The third patch fixes the warning to say "-Wc++11-compat" rather than
> "-Wc++0x-compat".
This also enables the following bogus warning:
~ % cat test.cpp
template struct X {};
template struct Y { static int const c = 0; };
int main() { return Y>::c; }
~ % g++ -Wall -std=c++11 test.cpp
test.cpp: In function ‘int main()’:
test.cpp:3:26: warning: ‘>>’ operator is treated as two right angle brackets in
C++11 [-Wc++11-compat]
int main() { return Y>::c; }
^
test.cpp:3:26: note: suggest parentheses around ‘>>’ expression
--
Markus
Re: PATCH] PR target/65612: Multiversioning doesn't work with DSO nor PIE
* H.J. Lu [2015-05-09 10:41:41 -0700]: > On Sat, May 9, 2015 at 7:31 AM, Szabolcs Nagy wrote: > > > > The symbol versioning hack for __cpu_model and __cpu_indicator_init > > makes them invisible to the musl dynamic linker so their relocation > > fails with 'symbol not found' error. > > (affects anything linked with -lgcc_s) > > There are > > 4: 2b70 806 FUNCGLOBAL DEFAULT 12 > __cpu_indicator_init@GCC_4.8.0 > 38: 002153d016 OBJECT GLOBAL DEFAULT 25 > __cpu_model@GCC_4.8.0 > > and > > 0215000 00040001 R_X86_64_64 > 2b70 __cpu_indicator_init@GCC_4.8.0 + 0 > 00215220 00260006 R_X86_64_GLOB_DAT > 002153d0 __cpu_model@GCC_4.8.0 + 0 > > in libgcc_s.so.1. Musl ld.so must be fixed to handle it. > (sorry, i missed the other discussion about the same issue) i still think a solution that avoids depending on symbol versioning in the dynamic linker would be preferable. (versioning is a sun/gnu elf extension as far as i understand, eg. not supported by posix dlsym). i think it might be enough to add __cpu_indicator_init_local as an alias to __cpu_indicator_init in libgcc.a and then use the *_local symbol from the ifunc resolver, that way no new dependency is added to libgcc_s.so handling. (adding versioning support to musl is problematic with the current code, a simpler workaround for musl is to provide its own weak definition of __cpu_indicator_init, but that's ugly.)
Re: PATCH] PR target/65612: Multiversioning doesn't work with DSO nor PIE
On Sat, May 09, 2015 at 10:41:41AM -0700, H.J. Lu wrote: > On Sat, May 9, 2015 at 7:31 AM, Szabolcs Nagy wrote: > > * H.J. Lu [2015-04-17 05:36:30 -0700]: > >> On Fri, Apr 17, 2015 at 4:59 AM, Jakub Jelinek wrote: > >> > On Fri, Apr 17, 2015 at 04:48:48AM -0700, H.J. Lu wrote: > >> >> > I don't like it. Nonshared libgcc is libgcc.a, period. No sense in > >> >> > creating yet another library for that. > >> >> > So, IMHO beyond making the __cpu* entrypoints compat symbols only (@ > >> >> > instead > >> >> > of @@ symbol versions) the right fix is simply tweak init_gcc_spec, > >> >> > so that > >> >> > static_name is always linked in, in the switch combinations that it > >> >> > isn't > >> >> > right now of course after shared_name rather than before that. > >> >> > I thought we've fixed that years ago... > >> >> > > > > > I think the patch committed for this is suboptimal. > > (it breaks with musl libc on x86 if libgcc_s is linked into a binary) > > > > the patch: > > http://gcc.gnu.org/ml/gcc-patches/2015-04/msg00878.html > > original thread: > > http://gcc.gnu.org/ml/gcc-patches/2015-03/msg01520.html > > > > The symbol versioning hack for __cpu_model and __cpu_indicator_init > > makes them invisible to the musl dynamic linker so their relocation > > fails with 'symbol not found' error. > > (affects anything linked with -lgcc_s) > > There are > > 4: 2b70 806 FUNCGLOBAL DEFAULT 12 > __cpu_indicator_init@GCC_4.8.0 > 38: 002153d016 OBJECT GLOBAL DEFAULT 25 > __cpu_model@GCC_4.8.0 > > and > > 0215000 00040001 R_X86_64_64 > 2b70 __cpu_indicator_init@GCC_4.8.0 + 0 > 00215220 00260006 R_X86_64_GLOB_DAT > 002153d0 __cpu_model@GCC_4.8.0 + 0 > > in libgcc_s.so.1. Musl ld.so must be fixed to handle it. The intent in musl was always not to support symbol versioning. There are various reasons for this, which I could go into, but I'd rather not turn this thread into an argument about the merits of symbol versioning. Originally, musl ignored the version data completely, and would happily have resolved the above symbol, but this also led to problems where third-party libraries used symbol versioning and at runtime we got the oldest-versioned symbol instead of the desired current one. So I changed the lookup to use the same logic as ld, rejecting all symbols with the hidden bit set, and getting us the newest one (which is the only one we intended to support). We can't just do the same version processing as glibc because we want symbols in libc itself to resolve regardless of the version in the reference (this is needed for the glibc ABI compat we provide). It might work to honor symbol versions only when the library being searched has versions, and otherwise allow symbols to match any version requested, but I haven't thought this through entirely yet. In any case I'd like the decision for what musl does about symbol versions (which are not intended to be supported, anyway) to be independent of libgcc's solution of this problem. We can solve it with a hack on our side (just providing dummy symbols by those names) but that's ugly too and I'd rather not do it. Could you clarify the reasoning for why libgcc is using this hack with a reference to an 'obsolete' symbol version rather than just visibility? Rich
Re: PATCH] PR target/65612: Multiversioning doesn't work with DSO nor PIE
On Sat, May 09, 2015 at 03:36:01PM -0400, Rich Felker wrote: > Could you clarify the reasoning for why libgcc is using this hack with > a reference to an 'obsolete' symbol version rather than just > visibility? Obviously for ABI compatibility reasons. Older programs could be relying on the symbols exported from libgcc_s.so.1 in the past. We need to keep those working, just don't want to allow linking new programs against this symbol (because new programs should get their copy from libgcc.a now). Jakub
Re: v3 PATCH to fix libstdc++ build/test with C++11 default compiler
On 05/08/2015 11:34 PM, Jason Merrill wrote:
The second patch fixes all but one of the testsuite failures that turned
up with a compiler that defaults to C++11.
And here's the fix for a C++14 compiler:
commit 42dc3f6f9af19c12d4201af8b6ef2e28bb08e944
Author: Jason Merrill
Date: Sat May 9 12:51:20 2015 -0500
* testsuite/20_util/headers/functional/synopsis.cc: Add
_GLIBCXX14_CONSTEXPR.
* testsuite/25_algorithms/headers/algorithm/synopsis.cc: Likewise.
diff --git a/libstdc++-v3/testsuite/20_util/headers/functional/synopsis.cc b/libstdc++-v3/testsuite/20_util/headers/functional/synopsis.cc
index 03d37a6..5e561de 100644
--- a/libstdc++-v3/testsuite/20_util/headers/functional/synopsis.cc
+++ b/libstdc++-v3/testsuite/20_util/headers/functional/synopsis.cc
@@ -50,9 +50,11 @@ namespace std {
// lib.negators, negators:
template struct unary_negate;
template
+ _GLIBCXX14_CONSTEXPR
unary_negate not1(const Predicate&);
template struct binary_negate;
template
+ _GLIBCXX14_CONSTEXPR
binary_negate not2(const Predicate&);
// lib.binders, binders:
diff --git a/libstdc++-v3/testsuite/25_algorithms/headers/algorithm/synopsis.cc b/libstdc++-v3/testsuite/25_algorithms/headers/algorithm/synopsis.cc
index ba38f25..4758796 100644
--- a/libstdc++-v3/testsuite/25_algorithms/headers/algorithm/synopsis.cc
+++ b/libstdc++-v3/testsuite/25_algorithms/headers/algorithm/synopsis.cc
@@ -483,75 +483,93 @@ namespace std
// 25.3.7, minimum and maximum:
template
+_GLIBCXX14_CONSTEXPR
const _Tp&
min(const _Tp&, const _Tp&);
template
+_GLIBCXX14_CONSTEXPR
const _Tp&
min(const _Tp&, const _Tp&, _Compare);
template
+_GLIBCXX14_CONSTEXPR
const _Tp&
max(const _Tp&, const _Tp&);
template
+_GLIBCXX14_CONSTEXPR
const _Tp&
max(const _Tp&, const _Tp&, _Compare);
template
+_GLIBCXX14_CONSTEXPR
_FIter
min_element(_FIter, _FIter);
template
+_GLIBCXX14_CONSTEXPR
_FIter
min_element(_FIter, _FIter, _Compare);
template
+_GLIBCXX14_CONSTEXPR
_FIter
max_element(_FIter, _FIter);
template
+_GLIBCXX14_CONSTEXPR
_FIter
max_element(_FIter, _FIter, _Compare);
#if __cplusplus >= 201103L
template
+_GLIBCXX14_CONSTEXPR
pair
minmax(const _Tp&, const _Tp&);
template
+_GLIBCXX14_CONSTEXPR
pair
minmax(const _Tp&, const _Tp&, _Compare);
template
+_GLIBCXX14_CONSTEXPR
pair<_FIter, _FIter>
minmax_element(_FIter, _FIter);
template
+_GLIBCXX14_CONSTEXPR
pair<_FIter, _FIter>
minmax_element(_FIter, _FIter, _Compare);
template
+_GLIBCXX14_CONSTEXPR
_Tp
min(initializer_list<_Tp>);
template
+_GLIBCXX14_CONSTEXPR
_Tp
min(initializer_list<_Tp>, _Compare);
template
+_GLIBCXX14_CONSTEXPR
_Tp
max(initializer_list<_Tp>);
template
+_GLIBCXX14_CONSTEXPR
_Tp
max(initializer_list<_Tp>, _Compare);
template
+_GLIBCXX14_CONSTEXPR
pair<_Tp, _Tp>
minmax(initializer_list<_Tp>);
template
+_GLIBCXX14_CONSTEXPR
pair<_Tp, _Tp>
minmax(initializer_list<_Tp>, _Compare);
#endif
Re: PATCHes to help with C++11 bootstrap
On 05/09/2015 01:27 PM, Markus Trippelsdorf wrote:
This also enables the following bogus warning:
~ % g++ -Wall -std=c++11 test.cpp
test.cpp:3:26: warning: ‘>>’ operator is treated as two right angle brackets in
C++11 [-Wc++11-compat]
Fixed thus:
commit 1c492cd18b2869305cc3ee16f84b6464f98f4e4c
Author: Jason Merrill
Date: Sat May 9 16:16:46 2015 -0500
* c-opts.c (c_common_post_options): Also clear
cpp_opts->cpp_warn_cxx11_compat.
diff --git a/gcc/c-family/c-opts.c b/gcc/c-family/c-opts.c
index bd99871..e9eb511 100644
--- a/gcc/c-family/c-opts.c
+++ b/gcc/c-family/c-opts.c
@@ -905,6 +905,7 @@ c_common_post_options (const char **pfilename)
/* If we're allowing C++0x constructs, don't warn about C++98
identifiers which are keywords in C++0x. */
warn_cxx11_compat = 0;
+ cpp_opts->cpp_warn_cxx11_compat = 0;
if (warn_narrowing == -1)
warn_narrowing = 1;
diff --git a/gcc/testsuite/g++.dg/cpp0x/bracket5.C b/gcc/testsuite/g++.dg/cpp0x/bracket5.C
new file mode 100644
index 000..22ac544
--- /dev/null
+++ b/gcc/testsuite/g++.dg/cpp0x/bracket5.C
@@ -0,0 +1,6 @@
+// { dg-options "-Wall" }
+// { dg-do compile { target c++11 } }
+
+template struct X {};
+template struct Y { static int const c = 0; };
+int main() { return Y>::c; }
Re: PATCHes to help with C++11 bootstrap
More patches:
The first makes changing the default just a matter of changing two lines
(in the compiler and testsuite).
The second patch is a minor tidy of c.opt.
Tested x86_64-pc-linux-gnu, applying to trunk.
commit 8c9891a3828dc6a4c91998b2437ef9fbf8659163
Author: Jason Merrill
Date: Fri May 8 17:01:37 2015 -0500
gcc/c-family/
* c-common.h (enum cxx_dialect): Add cxx_unset.
* c-common.c (cxx_dialect): Initialize to cxx_unset.
* c-opts.c (c_common_post_options): Set C++ dialect to C++98 if unset.
gcc/testsuite/
* lib/target-supports.exp (cxx_default): New global.
(check_effective_target_c++11_only)
(check_effective_target_c++14_only)
(check_effective_target_c++98_only)
(check_effective_target_c++1z_only): Check it.
diff --git a/gcc/c-family/c-common.c b/gcc/c-family/c-common.c
index 378f237..93b3060 100644
--- a/gcc/c-family/c-common.c
+++ b/gcc/c-family/c-common.c
@@ -256,9 +256,9 @@ const char *constant_string_class_name;
int flag_use_repository;
-/* The C++ dialect being used. C++98 is the default. */
+/* The C++ dialect being used. Default set in c_common_post_options. */
-enum cxx_dialect cxx_dialect = cxx98;
+enum cxx_dialect cxx_dialect = cxx_unset;
/* Maximum template instantiation depth. This limit exists to limit the
time it takes to notice excessively recursive template instantiations.
diff --git a/gcc/c-family/c-common.h b/gcc/c-family/c-common.h
index 603d3f0..5c1fa7b 100644
--- a/gcc/c-family/c-common.h
+++ b/gcc/c-family/c-common.h
@@ -671,6 +671,7 @@ extern int flag_use_repository;
/* The supported C++ dialects. */
enum cxx_dialect {
+ cxx_unset,
/* C++98 with TC1 */
cxx98,
cxx03 = cxx98,
diff --git a/gcc/c-family/c-opts.c b/gcc/c-family/c-opts.c
index 8b17674..bd99871 100644
--- a/gcc/c-family/c-opts.c
+++ b/gcc/c-family/c-opts.c
@@ -896,6 +896,10 @@ c_common_post_options (const char **pfilename)
if (flag_abi_version == 0)
flag_abi_version = 8;
+ /* Set C++ standard to C++98 if not specified on the command line. */
+ if (c_dialect_cxx () && cxx_dialect == cxx_unset)
+set_std_cxx98 (/*ISO*/false);
+
if (cxx_dialect >= cxx11)
{
/* If we're allowing C++0x constructs, don't warn about C++98
diff --git a/gcc/testsuite/lib/target-supports.exp b/gcc/testsuite/lib/target-supports.exp
index d68b48b..3728927 100644
--- a/gcc/testsuite/lib/target-supports.exp
+++ b/gcc/testsuite/lib/target-supports.exp
@@ -5892,15 +5892,23 @@ proc check_effective_target_c++ { } {
return 0
}
+set cxx_default "c++98"
# Check whether the current active language standard supports the features
-# of C++11/C++14 by checking for the presence of one of the -std
-# flags. This assumes that the default for the compiler is C++98, and that
+# of C++11/C++14 by checking for the presence of one of the -std flags.
+# This assumes that the default for the compiler is $cxx_default, and that
# there will never be multiple -std= arguments on the command line.
proc check_effective_target_c++11_only { } {
+global cxx_default
if ![check_effective_target_c++] {
return 0
}
-return [check-flags { { } { } { -std=c++0x -std=gnu++0x -std=c++11 -std=gnu++11 } }]
+if [check-flags { { } { } { -std=c++0x -std=gnu++0x -std=c++11 -std=gnu++11 } }] {
+ return 1
+}
+if { $cxx_default == "c++11" && [check-flags { { } { } { } { -std=* } }] } {
+ return 1
+}
+return 0
}
proc check_effective_target_c++11 { } {
if [check_effective_target_c++11_only] {
@@ -5912,14 +5920,21 @@ proc check_effective_target_c++11_down { } {
if ![check_effective_target_c++] {
return 0
}
-return ![check_effective_target_c++14]
+return [expr ![check_effective_target_c++14] ]
}
proc check_effective_target_c++14_only { } {
+global cxx_default
if ![check_effective_target_c++] {
return 0
}
-return [check-flags { { } { } { -std=c++14 -std=gnu++14 -std=c++14 -std=gnu++14 } }]
+if [check-flags { { } { } { -std=c++14 -std=gnu++14 -std=c++14 -std=gnu++14 } }] {
+ return 1
+}
+if { $cxx_default == "c++14" && [check-flags { { } { } { } { -std=* } }] } {
+ return 1
+}
+return 0
}
proc check_effective_target_c++14 { } {
@@ -5932,21 +5947,35 @@ proc check_effective_target_c++14_down { } {
if ![check_effective_target_c++] {
return 0
}
-return ![check_effective_target_c++1z]
+return [expr ![check_effective_target_c++1z] ]
}
proc check_effective_target_c++98_only { } {
+global cxx_default
if ![check_effective_target_c++] {
return 0
}
-return ![check_effective_target_c++11]
+if [check-flags { { } { } { -std=c++98 -std=gnu++98 -std=c++03 -std=gnu++03 } }] {
+ return 1
+}
+if { $cxx_default == "c++98" && [check-flags { { } { } { } { -std=* } }] } {
+ return 1
+}
+return 0
}
proc check_effective_target_c++1z_only { } {
+global cxx_default
if ![check_effective_target_c++] {
return
Re: [patch, Fortran] Fix PR 66041
Am 09.05.2015 um 13:59 schrieb Mikael Morin:
> You also need to remove/free the trailing subreferences.
That's right, I did that. Although I will probably never understand
why lbound(a) should be different from lbound(a%r)...
>> + /* We have to get rid of the shape, if thre is one. Do
>> + so by freeing it and calling gfc_resolve to rebuild it,
>> + if necessary. */
>> +
>> + if (lbound_e->shape)
>> +gfc_free_shape (&(lbound_e->shape), lbound_e->rank);
>> +
>
>> + lbound_e->rank = ar->dimen;
> ar->dimen is not what you think it is.
> It is 3 for array(1, 1, :), while the rank is 1.
> gfc_resolve_expr should set the rank for you, so just remove this line.
It doesn't (for whatever reason), so I kept on setting it.
>> +
>> + gfc_resolve_expr (lbound_e);
>> + lbound = get_array_inq_function (GFC_ISYM_LBOUND,
>> + lbound_e, i + 1);
> free lbound_e?
It will be part of the lbound expression, or be simplified away.
>
>> }
>> - lbound = get_array_inq_function (GFC_ISYM_LBOUND, e_in,
>> - i_index + 1);
>> + else
>> +lbound = get_array_inq_function (GFC_ISYM_LBOUND, e_in,
>> + i_index + 1);
> You can't reuse e_in if it has subreferences.
Changed.
>> }
>>
>>ar->dimen_type[i] = DIMEN_ELEMENT;
>> @@ -2639,6 +2665,8 @@ scalarized_expr (gfc_expr *e_in, gfc_expr **index,
>>i_index ++;
>> }
>> }
>> + gfc_free_expr (e_in);
>> +
> This side effect is asking for trouble.
> Instead of this, remove the copies made in the callers.
> This is independant from the rest, so it can be made later as a follow-up.
Done (all in once).
I have attached the new patch (in which I also restructured the test),
plus the test cases.
OK for trunk?
Thomas
2015-05-08 Thomas Koenig
PR fortran/66041
* frontend-passes.c (scalarized_expr): Copy first argument so it
is not necessary to call gfc_copy_expr() on its argument. Set
correct dimension and shape for the expression to be passed to
lbound. Remove trailing references after array refrence.
(inline_matmul_assign): Remove gfc_copy_expr() from calls
to scalarized_expr().
2015-05-08 Thomas Koenig
PR fortran/66041
* gfortran.dg/inline_matmul_7.f90: New test.
* gfortran.dg/inline_matmul_8.f90: New test.
* gfortran.dg/inline_matmul_9.f90: New test.
Index: frontend-passes.c
===
--- frontend-passes.c (Revision 222864)
+++ frontend-passes.c (Arbeitskopie)
@@ -2532,16 +2532,17 @@ get_size_m1 (gfc_expr *e, int dimen)
references have been frozen. */
static gfc_expr*
-scalarized_expr (gfc_expr *e_in, gfc_expr **index, int count_index)
+scalarized_expr (gfc_expr *ei, gfc_expr **index, int count_index)
{
gfc_array_ref *ar;
int i;
int rank;
- gfc_expr *e;
+ gfc_expr *e, *e_in;
int i_index;
bool was_fullref;
- e = gfc_copy_expr(e_in);
+ e = gfc_copy_expr(ei);
+ e_in = gfc_copy_expr (ei);
rank = e->rank;
@@ -2607,18 +2608,54 @@ static gfc_expr*
}
else
{
+ gfc_expr *lbound_e;
+ gfc_ref *ref;
+
+ lbound_e = gfc_copy_expr (e_in);
+
+ for (ref = lbound_e->ref; ref; ref = ref->next)
+ if (ref->type == REF_ARRAY
+ && (ref->u.ar.type == AR_FULL
+ || ref->u.ar.type == AR_SECTION))
+ break;
+
+ if (ref->next)
+ {
+ gfc_free_ref_list (ref->next);
+ ref->next = NULL;
+ }
+
if (!was_fullref)
{
/* Look at full individual sections, like a(:). The first index
is the lbound of a full ref. */
-
+ int j;
gfc_array_ref *ar;
- ar = gfc_find_array_ref (e_in);
+ ar = &ref->u.ar;
ar->type = AR_FULL;
+ for (j = 0; j < ar->dimen; j++)
+ {
+ gfc_free_expr (ar->start[j]);
+ ar->start[j] = NULL;
+ gfc_free_expr (ar->end[j]);
+ ar->end[j] = NULL;
+ gfc_free_expr (ar->stride[j]);
+ ar->stride[j] = NULL;
+ }
+
+ /* We have to get rid of the shape, if there is one. Do
+ so by freeing it and calling gfc_resolve to rebuild
+ it, if necessary. */
+
+ if (lbound_e->shape)
+ gfc_free_shape (&(lbound_e->shape), lbound_e->rank);
+
+ lbound_e->rank = ar->dimen;
+ gfc_resolve_expr (lbound_e);
}
- lbound = get_array_inq_function (GFC_ISYM_LBOUND, e_in,
- i_index + 1);
+ lbound = get_array_inq_function (GFC_ISYM_LBOUND, lbound_e,
+ i + 1);
}
ar->dimen_type[i] = DIMEN_ELEMENT;
@@ -2639,6 +2676,8 @@ static gfc_expr*
i_index ++;
}
Re: PATCHes to help with C++11 bootstrap
On 05/09/2015 05:37 AM, Richard Biener wrote: Hmm, I wonder if we want to bootstrap with explicit -std=gnu04, our host compiler requirement. Otherwise we'll silently sneak in C++11 features when that becomes the default? I think just for stage 1. Jason
