[Bug c/33111] New: Bad code generation with -O2 (ARM 7 architecture)
#define UIP_LLH_LEN 1
#define BUF ((uip_tcpip_hdr *)uip_buf[UIP_LLH_LEN])
#define u8_t unsigned char
#define u16_t unsigned short
#define UIP_BUFSIZE 1000
typedef struct {
/* IP header. */
u8_t vhl,
tos,
len[2],
ipid[2],
ipoffset[2],
ttl,
proto;
u16_t ipchksum;
u16_t srcipaddr[2],
destipaddr[2];
/* TCP header. */
u16_t srcport,
destport;
u8_t seqno[4],
ackno[4],
tcpoffset,
flags,
wnd[2];
u16_t tcpchksum;
u8_t urgp[2];
u8_t optdata[4];
} uip_tcpip_hdr;
u8_t uip_buf[UIP_BUFSIZE+2];
volatile u8_t uip_acc32[4];
int jaroslav;
int main(void)
{
if((BUF-ackno[0] == uip_acc32[0])
(BUF-ackno[1] == uip_acc32[1])
(BUF-ackno[2] == uip_acc32[2])
(BUF-ackno[3] == uip_acc32[3]))
jaroslav++;
return 1;
}
.file main.c
.text
.align 2
.global main
.type main, %function
main:
@ args = 0, pretend = 0, frame = 0
@ frame_needed = 0, uses_anonymous_args = 0
@ link register save eliminated.
ldr r0, .L10
ldr r1, .L10+4
ldrbr2, [r0, #28] @ zero_extendqisi2
ldrbr3, [r1, #0]@ zero_extendqisi2
cmp r2, r3
@ lr needed for prologue
beq .L9
.L2:
mov r0, #1
bx lr
.L9:
ldrbr2, [r1, #1]@ zero_extendqisi2
ldrbr3, [r0, #29] @ zero_extendqisi2
cmp r3, r2
bne .L2
ldrbr2, [r1, #2]@ zero_extendqisi2
ldrbr3, [r0, #30] @ zero_extendqisi2
cmp r3, r2
bne .L2
ldrbr2, [r1, #3]@ zero_extendqisi2
ldrbr3, [r0, #31] @ zero_extendqisi2
cmp r3, r2 I THINK CC's OF THIS
COMPARE NEVER GET LOOKED AT
ldreq r2, .L10+8
ldreq r3, [r2, #0]
addeq r3, r3, #1
streq r3, [r2, #0]
b .L2
.L11:
.align 2
.L10:
.word uip_buf+1
.word uip_acc32
.word jaroslav
.size main, .-main
.comm uip_buf,1002,1
.comm uip_acc32,4,1
.comm jaroslav,4,4
.ident GCC: (GNU) 4.2.0
--
Summary: Bad code generation with -O2 (ARM 7 architecture)
Product: gcc
Version: 4.2.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: gressau at optusnet dot com dot au
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33111
[Bug c/33111] Bad code generation with -O2 (ARM 7 architecture)
--- Comment #1 from pinskia at gcc dot gnu dot org 2007-08-19 08:59 --- #define BUF ((uip_tcpip_hdr *)uip_buf[UIP_LLH_LEN]) Right there is an alias violation. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33111
[Bug rtl-optimization/32557] [4.3 Regression] internal compiler error: RTL check: expected code 'reg', have 'subreg' in rhs_regno, at rtl.h:956
--- Comment #4 from patchapp at dberlin dot org 2007-08-19 12:30 --- Subject: Bug number PR rtl-optimization/32557 A patch for this bug has been added to the patch tracker. The mailing list url for the patch is http://gcc.gnu.org/ml/gcc-patches/2007-08/msg01210.html -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32557
[Bug c++/33112] New: ICE : canonical types differ for identical types const _CharT* [14] and const _CharT* [14]
This is with today's g++ 4.3, but has been there for a few weeks: Configured with: /proj/geometrica/home/GCC/trunk/configure --with-gmp=/proj/geometrica/home/MPFR/MPFR-2.2.1/FC5 --disable-shared --disable-nls --enable-languages=c++,fortran --prefix=/proj/geometrica/home/GCC/Linux-fc5-4.3 Thread model: posix gcc version 4.3.0 20070819 (experimental) Compiling the attached preprocessed file with -c -O2 produces the following ICE. Note that it is not deterministic: you usually need to repeat the compile 10-20 times to see it... [EMAIL PROTECTED]:/proj/geometrica/home/CGAL/CGAL-I/src/CGAL while (/proj/geometrica/home/GCC/Linux-fc5-4.3/bin/g++ -c JAMA_numeric_solver.ii -O2 ) ; do ; echo OK ;done; OK OK OK OK OK In file included from /auto/papillon/u/papillon/0/projserver/geometrica/home/GCC/Linux-fc5-4.3/bin/../lib/gcc/i686-pc-linux-gnu/4.3.0/../../../../include/c++/4.3.0/locale:49, from /proj/geometrica/home/CGAL/CGAL-I/include/CGAL/Quotient.h:41, from /proj/geometrica/home/CGAL/CGAL-I/include/CGAL/Root_of_traits.h:27, from /proj/geometrica/home/CGAL/CGAL-I/include/CGAL/number_type_basic.h:87, from /proj/geometrica/home/CGAL/CGAL-I/include/CGAL/basic.h:36, from /proj/geometrica/home/CGAL/CGAL-I/include/CGAL/Polynomial/internal/macros.h:31, from /proj/geometrica/home/CGAL/CGAL-I/include/CGAL/Polynomial/basic.h:42, from /proj/geometrica/home/CGAL/CGAL-I/include/CGAL/Polynomial/internal/numeric_solvers_support.h:24, from JAMA_numeric_solver.cpp:22: /auto/papillon/u/papillon/0/projserver/geometrica/home/GCC/Linux-fc5-4.3/bin/../lib/gcc/i686-pc-linux-gnu/4.3.0/../../../../include/c++/4.3.0/bits/locale_facets_nonio.h:176: internal compiler error: canonical types differ for identical types const _CharT* [14] and const _CharT* [14] Please submit a full bug report, with preprocessed source if appropriate. See URL:http://gcc.gnu.org/bugs.html for instructions. I have also been able to reproduce it with only -O. -- Summary: ICE : canonical types differ for identical types const _CharT* [14] and const _CharT* [14] Product: gcc Version: 4.3.0 Status: UNCONFIRMED Severity: normal Priority: P3 Component: c++ AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: sylvain dot pion at sophia dot inria dot fr GCC host triplet: i686-pc-linux-gnu http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33112
[Bug c++/33112] ICE : canonical types differ for identical types const _CharT* [14] and const _CharT* [14]
--- Comment #1 from sylvain dot pion at sophia dot inria dot fr 2007-08-19 12:55 --- Created an attachment (id=14076) -- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=14076action=view) Compressed pre-processed C++ source code -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33112
[Bug target/30315] optimize unsigned-add overflow test on x86 to use cpu flags from addl
--- Comment #12 from rask at gcc dot gnu dot org 2007-08-19 13:00 ---
void *foo(unsigned a, unsigned b) {
unsigned sum = a + b;
if (sum a) return 0;
if (sum == 0) sum = 1;
To be able to optimize both comparisons, we need to have the same comparison
operands. The only way I can see this happening is to canonicalize the overflow
check as a + b 0 with the obvious fear that since a and b are both
unsigned, GCC will optimize the comparison away as always false - unless taught
not to do so. I'll see what I can do.
Additionally, the resulting asm seems to be a bit stupid:
testl %edx, %edx
movl$1, %eax
cmove %eax, %edx
should be just
testl %edx, %edx
sete%dl
--
rask at gcc dot gnu dot org changed:
What|Removed |Added
AssignedTo|unassigned at gcc dot gnu |rask at gcc dot gnu dot org
|dot org |
Status|UNCONFIRMED |ASSIGNED
Ever Confirmed|0 |1
Last reconfirmed|-00-00 00:00:00 |2007-08-19 13:00:47
date||
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30315
[Bug c++/33112] ICE : canonical types differ for identical types const _CharT* [14] and const _CharT* [14]
-- steven at gcc dot gnu dot org changed: What|Removed |Added CC||dgregor at gcc dot gnu dot ||org Status|UNCONFIRMED |NEW Ever Confirmed|0 |1 Last reconfirmed|-00-00 00:00:00 |2007-08-19 13:04:48 date|| http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33112
[Bug c/33111] Bad code generation with -O2 (ARM 7 architecture)
--- Comment #2 from gressau at optusnet dot com dot au 2007-08-19 13:23 --- The attached sample code does not illustrate the real problem. Further investigation is required why the code works with -O0 and does not work with -O2. -- gressau at optusnet dot com dot au changed: What|Removed |Added CC|gressau at optusnet dot com | |dot au | http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33111
[Bug tree-optimization/33113] New: Failing to represent the stride of a dataref when it is not a constant
In the following testcase: subroutine sub(aa,bb,n,m) implicit none integer, intent(in) :: n,m real, intent(inout) :: aa(n,m) real, intent(in):: bb(n,m) integer :: i,j do i = 1,m do j= 2,n aa(i,j)= aa(i,j-1)+bb(i,j-1) enddo enddo end subroutine end The stride of the accesses in the inner-loop is a parameter (m is not a compile-time known constant). As a result the data dataref analyzer reports: failed: evolution of offset is not affine ... base_address: offset from base address: constant offset from base address: step: aligned to: base_object: (*aa_54(D))[0] symbol tag: SMT.25 Any chance that the dataref analysis can return an (invariant) expression in step, so that further analysis could continue? (for example, the access in the outer-loop is consecutive, so if we had an expression to represent the inner-loop stride, we could vectorize the outer-loop). -- Summary: Failing to represent the stride of a dataref when it is not a constant Product: gcc Version: 4.3.0 Status: UNCONFIRMED Keywords: missed-optimization Severity: normal Priority: P3 Component: tree-optimization AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: dorit at gcc dot gnu dot org GCC build triplet: powerpc64-linux GCC host triplet: powerpc64-linux GCC target triplet: powerpc64-linux http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33113
[Bug tree-optimization/32378] can't determine dependence (distinct sections of an array)
--- Comment #6 from dorit at gcc dot gnu dot org 2007-08-19 13:47 ---
Sebastian - any thughts/plans?
Here's another testcase:
subroutine sub(aa,bb,n,m)
implicit none
integer, intent(in) :: n,m
real, intent(inout) :: aa(n,m)
real, intent(in):: bb(n,m)
integer :: i,j
do j= 2,n
do i = 1,m
aa(i,j)= aa(i,j-1)+bb(i,j-1)
enddo
enddo
end subroutine
end
Here too we get:
(compute_affine_dependence
(stmt_a =
D.1385_55 = (*aa_54(D))[D.1384_53])
(stmt_b =
(*aa_54(D))[D.1380_49] = D.1390_62)
(subscript_dependence_tester
(analyze_overlapping_iterations
(chrec_a = {pretmp.34_76 + 1, +, 1}_2)
(chrec_b = {pretmp.34_32 + 1, +, 1}_2)
(analyze_siv_subscript
siv test failed: unimplemented.
)
(overlap_iterations_a = not known
)
(overlap_iterations_b = not known
)
)
(dependence classified: scev_not_known)
)
)
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32378
[Bug tree-optimization/33114] New: jc1 segault building lljava/java/util/AbstractMap.java
Using gcc 4.3.0 20070818 (experimental) has tI hit the following problem building libjava libtool: compile: /home/dps/src/egcs/gcc-i686/gcc/gcj -B/home/dps/src/egcs/gcc-i686/i686-pc-linux-gnu/libjava/ -B/home/dps/src/egcs/gcc-i686/gcc/ -ffloat-store -fomit-frame-pointer -Usun -fclasspath= -fbootclasspath=../../../gcc/libjava/classpath/lib --encoding=UTF-8 -Wno-deprecated -fbootstrap-classes -g -O2 -c -fsource-filename=/home/dps/src/egcs/gcc-i686/i686-pc-linux-gnu/libjava/classpath/lib/classes -MT java/util.lo -MD -MP -MF java/util.deps @java/util.list -fPIC -o java/.libs/util.o ../../../../../gcc/libjava/classpath/java/util/AbstractMap.java: In class 'java.util.AbstractMap$2$1': ../../../../../gcc/libjava/classpath/java/util/AbstractMap.java: In constructor '(java.util.AbstractMap$2)': ../../../../../gcc/libjava/classpath/java/util/AbstractMap.java:604: error: class 'java.util.AbstractMap$2' has no method named 'access$0' matching signature '(Ljava/util/AbstractMap$2;)Ljava/util/AbstractMap;' ../../../../../gcc/libjava/classpath/java/util/AbstractMap.java: In class 'java.util.AbstractMap$3': ../../../../../gcc/libjava/classpath/java/util/AbstractMap.java: In method 'access$0(java.util.AbstractMap$3)': ../../../../../gcc/libjava/classpath/java/util/AbstractMap.java:0: internal compiler error: Segmentation fault Apprerently the reason for this that fini_ssa_operands is called more times than init_ssa_operands. Backtrace of call with n_initialized=0 is (gdb) bt #0 fini_ssa_operands () at ../../gcc/gcc/tree-ssa-operands.c:389 #1 0x08403a40 in delete_tree_ssa () at ../../gcc/gcc/tree-ssa.c:866 #2 0x0832fedc in execute_free_datastructures () at ../../gcc/gcc/tree-optimize.c:215 #3 0x0824b79f in execute_one_pass (pass=0x87c3540) at ../../gcc/gcc/passes.c:1108 #4 0x0824b99f in execute_pass_list (pass=0x87c3540) at ../../gcc/gcc/passes.c:1161 #5 0x0832ffed in tree_rest_of_compilation (fndecl=0xb7c98f80) at ../../gcc/gcc/tree-optimize.c:405 #6 0x084a5cf0 in cgraph_expand_function (node=0xb7454280) at ../../gcc/gcc/cgraphunit.c:1077 #7 0x084a7060 in cgraph_assemble_pending_functions () at ../../gcc/gcc/cgraphunit.c:440 #8 0x084a74ed in cgraph_finalize_function (decl=0xb7c98f80, nested=0 '\0') at ../../gcc/gcc/cgraphunit.c:557 #9 0x08064889 in finish_method (fndecl=0xb7c98f80) at ../../gcc/gcc/java/decl.c:1859 #10 0x08070415 in end_java_method () at ../../gcc/gcc/java/decl.c:1807 #11 0x08095308 in parse_class_file () at ../../gcc/gcc/java/jcf-parse.c:1689 #12 0x08095cf3 in java_parse_file (set_yydebug=0) at ../../gcc/gcc/java/jcf-parse.c:1980 #13 0x082ce838 in toplev_main (argc=34, argv=0xbfdc8ba4) at ../../gcc/gcc/toplev.c:1043 #14 0x080a15bf in main (argc=Cannot access memory at address 0x4 ) at ../../gcc/gcc/main.c:35 n_initialized reaches -4 and then the code segaults on line 1092 of upodate_stmt_operands. Backtrace is #0 0x083c46b3 in update_stmt_operands (stmt=0xb7469c40) at ../../gcc/gcc/tree-ssa-operands.c:1092 #1 0x082d3742 in bsi_insert_after (i=0xbfdc8780, t=0xb7469c40, m=BSI_NEW_STMT) at ../../gcc/gcc/tree-flow-inline.h:368 #2 0x084c83ff in copy_body (id=0xbfdc881c, count=value optimized out, frequency=0, entry_block_map=0xb746b8e8, exit_block_map=0xb746b924) at ../../gcc/gcc/tree-inline.c:811 #3 0x084c9b8d in tree_function_versioning (old_decl=0xb7ca0200, new_decl=0xb7467680, tree_map=0x0, update_clones=1 '\001') at ../../gcc/gcc/tree-inline.c:3445 #4 0x084a6b97 in save_inline_function_body (node=0xb7467380) at ../../gcc/gcc/cgraphunit.c:1626 #5 0x084ac22e in apply_inline () at ../../gcc/gcc/ipa-inline.c:1580 #6 0x0824b79f in execute_one_pass (pass=0x87c4ca0) at ../../gcc/gcc/passes.c:1108 #7 0x0824b99f in execute_pass_list (pass=0x87c4ca0) at ../../gcc/gcc/passes.c:1161 #8 0x0832ffed in tree_rest_of_compilation (fndecl=0xb7ca0200) at ../../gcc/gcc/tree-optimize.c:405 #9 0x084a5cf0 in cgraph_expand_function (node=0xb7467380) at ../../gcc/gcc/cgraphunit.c:1077 #10 0x084a7060 in cgraph_assemble_pending_functions () at ../../gcc/gcc/cgraphunit.c:440 #11 0x084a74ed in cgraph_finalize_function (decl=0xb7ca0200, nested=0 '\0') at ../../gcc/gcc/cgraphunit.c:557 #12 0x08064889 in finish_method (fndecl=0xb7ca0200) at ../../gcc/gcc/java/decl.c:1859 #13 0x08070415 in end_java_method () at ../../gcc/gcc/java/decl.c:1807 #14 0x08095308 in parse_class_file () at ../../gcc/gcc/java/jcf-parse.c:1689 #15 0x08095cf3 in java_parse_file (set_yydebug=0) at ../../gcc/gcc/java/jcf-parse.c:1980 #16 0x082ce838 in toplev_main (argc=34, argv=0xbfdc8ba4) at ../../gcc/gcc/toplev.c:1043 #17 0x080a15bf in main (argc=65715, argv=0x0) at ../../gcc/gcc/main.c:35 -- Summary: jc1 segault building lljava/java/util/AbstractMap.java Product: gcc Version: 4.3.0 Status: UNCONFIRMED Severity: normal Priority: P3 Component: tree-optimization AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: dps at simpson dot demon dot
[Bug preprocessor/30786] [4.1/4.2/4.3 Regression] ICE on _Pragma at end of file
--- Comment #4 from tromey at gcc dot gnu dot org 2007-08-19 16:13 --- I'm testing this patch. -- tromey at gcc dot gnu dot org changed: What|Removed |Added AssignedTo|unassigned at gcc dot gnu |tromey at gcc dot gnu dot |dot org |org Status|NEW |ASSIGNED Last reconfirmed|2007-02-13 15:55:12 |2007-08-19 16:13:10 date|| http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30786
[Bug fortran/33106] Access of components of public entities of private types wrongly allowed
--- Comment #1 from dfranke at gcc dot gnu dot org 2007-08-19 17:05 --- Confirmed and taken. -- dfranke at gcc dot gnu dot org changed: What|Removed |Added CC||dfranke at gcc dot gnu dot ||org AssignedTo|unassigned at gcc dot gnu |dfranke at gcc dot gnu dot |dot org |org Status|UNCONFIRMED |ASSIGNED Ever Confirmed|0 |1 Last reconfirmed|-00-00 00:00:00 |2007-08-19 17:05:50 date|| http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33106
[Bug fortran/33106] Access of components of public entities of private types wrongly allowed
--- Comment #2 from dfranke at gcc dot gnu dot org 2007-08-19 17:40 --- The print statements are wrong(?); additionally for -std=f95 there should be already an error for the compilation of the module. Tobias, does any other standard but F95 allow the usage of private derived type outside their defining module? I'd think that an error is called for, regardless of the version of the standard?! -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33106
[Bug fortran/33106] Access of components of public entities of private types wrongly allowed
--- Comment #3 from burnus at gcc dot gnu dot org 2007-08-19 17:50 ---
does any other standard but F95 allow the usage of private derived type
outside their defining module? I'd think that an error is called for,
regardless of the version of the standard?!
Believing NAG the following is allowed in Fortran 2003 but not in Fortran 95:
module m
type, private :: hidden_type; character(6) :: code; end type
type(hidden_type), public, parameter :: code_green = hidden_type('green')
end module
use m, only: code_green, module_procedure_of_m
call module_procedure_of_m(code_green)
use m, only: hidden_type is definitely invalid. I would expect that also
print *, code_green and print *, code_green%code is invalid.
According to NAG is for Fortran 95 type(hidden_type), public already wrong.
I will try to find the relevant part in the Fortran 95 and 2003 standard (is
probably a bit hidden).
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33106
[Bug target/30315] optimize unsigned-add overflow test on x86 to use cpu flags from addl
--- Comment #13 from ubizjak at gmail dot com 2007-08-19 18:03 --- (In reply to comment #12) if (sum == 0) sum = 1; Additionally, the resulting asm seems to be a bit stupid: testl %edx, %edx movl$1, %eax cmove %eax, %edx should be just testl %edx, %edx sete%dl Not, because sete sets dl to 0 (zero) when condition is not met. The c code above doesn't change sum when sum != 0. Regarding the problem, described in comment #11: Perhaps we can add CCCZmode as a common mode to otherwise orthogonal CCCmode and CCZmode. Playing with i386_cc_modes_compatible() and ix86_match_cc_mode() we can perhaps teach combine to CSE and merge two comparisons. To achieve this, I think we should use (enhanced) ix86_match_cc_mode() in just added RTL patterns. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30315
[Bug preprocessor/30786] [4.1/4.2/4.3 Regression] ICE on _Pragma at end of file
--- Comment #5 from tromey at gcc dot gnu dot org 2007-08-19 18:19 --- A program like '_Pragma /*comment*/' still gives an ICE. I'm testing an updated patch. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30786
[Bug fortran/33106] Access of components of public entities of private types wrongly allowed
--- Comment #4 from burnus at gcc dot gnu dot org 2007-08-19 18:54 --- Fortran 2003: --- 4.5.1.1 Accessibility Types that are defined in a module or accessible in that module by use association have either the PUBLIC or PRIVATE attribute. Types for which an access-spec is not explicitly specified in that module have the default accessibility attribute for that module. The default accessibility attribute for a module is PUBLIC unless it has been changed by a PRIVATE statement (5.2.1). Only types that have the PUBLIC attribute in that module are available to be accessed from that module by use association. The accessibility of a type does not affect, and is not affected by, the accessibility of its components and bindings. If a type definition is private, then the type name, and thus the structure constructor (4.5.9) for the type, are accessible only within the module containing the definition. --- Fortran 95: --- 4.4.1 Derived-type definition [...] The accessibility of a derived type may be declared explicitly by an access-spec in its derived-type-stmt or in an access-stmt (5.2.3). The accessibility is the default if it is not declared explicitly. If a type definition is private, then the type name, the structure constructor (4.4.4) for the type, any entity that is of the type, and any procedure that has a dummy argument or function result that is of the type are accessible only within the module containing the definition. If a type definition contains a PRIVATE statement, the component names for the type are accessible only within the module containing the definition, even if the type itself is public (5.1.2.2). The component names and hence the internal structure of the type are inaccessible in any scoping unit accessing the module via a USE statement. Similarly, the structure constructor for such a type shall be employed only within the defining module. --- Fortran 95 prohibits: If a type definition is private, then [...] any entity that is of the type, and any procedure that has a dummy argument or function result that is of the type are accessible only within the module. But Fortran 2003 only prohibits the type name and structure constructor. Thus entities and procedures with dummy arguments of this type are allowed. I would also argue that in Fortran 2003: print *, code_green%code is allowed as The accessibility of a type does not affect, and is not affected by, the accessibility of its components. Which agrees with NAG f95. For print *, code_green I am not fully sure, but one may argue that the type definition of a PRIVATE type (name) is PRIVATE and thus also the type spec of code_green. This seems to be the line NAG f95 follows. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33106
[Bug preprocessor/30786] [4.1/4.2/4.3 Regression] ICE on _Pragma at end of file
--- Comment #6 from tromey at gcc dot gnu dot org 2007-08-19 18:54 --- What does a plain _Pragma mean? Currently we try to give an error for this. However, it is not clear that this is actually erroneous. The copy of the standard that I am reading only mentions: A unary operator expression of the form: _Pragma ( string-literal ) ... but a plain _Pragma does not have that form. I'm not even sure that _Pragma(something-other-than-string-literal) should be recognized (but I'm new to C standardese...) Perhaps I have the wrong version of the standard or perhaps there is some more recent development of which I'm unaware. Anyway if we assume that a plain _Pragma is not an error then the first patch here is close to the correct one -- it just returns the plain _Pragma to the FE as an identifier. If all forms other than the full form are unrecognized then I think this first patch is correct. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30786
[Bug tree-optimization/30564] [4.3 Regression] ice for legal code with -O3
--- Comment #14 from pinskia at gcc dot gnu dot org 2007-08-19 19:07
---
The testcases here don't crash anymore but the one from PR 32033 does:
static int spready[] = {
};
void explosion_map (int y)
{
for (int i = 0; i 4; i++)
if (y * spready[i] 0)
break;
}
void explosion (void)
{
explosion_map (0);
for (int i = 0; i 2; i++)
continue;
}
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30564
[Bug tree-optimization/30564] [4.3 Regression] ice for legal code with -O3
--- Comment #15 from pinskia at gcc dot gnu dot org 2007-08-19 19:19 --- Testing a slight modifed version of the patch. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30564
[Bug libfortran/30162] I/O with named pipes does not work
--- Comment #21 from jvdelisle at gcc dot gnu dot org 2007-08-19 19:45 --- This works on 4.2 and 4.3. Not planning to backport to 4.1. See if you can get a later version of gfortran. If you have problems doing so, check the gfortran wiki binaries. If that fails, let me know. -- jvdelisle at gcc dot gnu dot org changed: What|Removed |Added Status|NEW |RESOLVED Resolution||WONTFIX http://gcc.gnu.org/bugzilla/show_bug.cgi?id=30162
[Bug target/33115] New: -march=native is not supported under x86 darwin
Mentioned here: http://gcc.gnu.org/ml/gcc-help/2007-08/msg00200.html And I tested it to make sure it was still not supported: [andrew-pinskis-computer:~] apinski% ~/local-gcc/bin/gcc -march=native t.c -O2 -o - -S t.c:1: error: bad value (native) for -march= switch t.c:1: error: bad value (native) for -mtune= switch [andrew-pinskis-computer:~] apinski% uname -a Darwin andrew-pinskis-computer.local 8.10.1 Darwin Kernel Version 8.10.1: Wed May 23 16:33:00 PDT 2007; root:xnu-792.22.5~1/RELEASE_I386 i386 i386 -- Summary: -march=native is not supported under x86 darwin Product: gcc Version: 4.3.0 Status: UNCONFIRMED Severity: enhancement Priority: P3 Component: target AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: pinskia at gcc dot gnu dot org GCC build triplet: i386-apple-darwin8.10 GCC host triplet: i386-apple-darwin8.10 GCC target triplet: i386-apple-darwin8.10 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33115
[Bug middle-end/33029] [4.3 Regression] libgcc2.c:1890: internal compiler error: in local_cprop_pass, at gcse.c:3236
--- Comment #8 from andreast at gcc dot gnu dot org 2007-08-19 19:56 --- I can confirm the patch in comment 7 fixes the bootstrap failure here. Thanks! -- andreast at gcc dot gnu dot org changed: What|Removed |Added CC||andreast at gcc dot gnu dot ||org http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33029
[Bug tree-optimization/33113] Failing to represent the stride (with array) of a dataref when it is not a constant
--- Comment #1 from rakdver at gcc dot gnu dot org 2007-08-19 20:08 --- (In reply to comment #0) In the following testcase: subroutine sub(aa,bb,n,m) implicit none integer, intent(in) :: n,m real, intent(inout) :: aa(n,m) real, intent(in):: bb(n,m) integer :: i,j do i = 1,m do j= 2,n aa(i,j)= aa(i,j-1)+bb(i,j-1) enddo enddo end subroutine end The stride of the accesses in the inner-loop is a parameter (m is not a compile-time known constant). As a result the data dataref analyzer reports: failed: evolution of offset is not affine ... base_address: offset from base address: constant offset from base address: step: aligned to: base_object: (*aa_54(D))[0] symbol tag: SMT.25 Any chance that the dataref analysis can return an (invariant) expression in step, so that further analysis could continue? (for example, the access in the outer-loop is consecutive, so if we had an expression to represent the inner-loop stride, we could vectorize the outer-loop). Making us return symbolic stride would not be hard. The problem is that data dependence analysis would fail anyway, since we cannot tell whether n is zero. -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33113
[Bug fortran/33116] New: External functions can be called as subroutine
If one uses EXTERNAL foo ! or PROCEDURE() :: foo foo might be either function or a subroutine. However, using REAL :: foo EXTERNAL foo ! or: PROCEDURE(REAL) :: foo foo is a function. However, gfortran happily allows the following: EXTERNAL foo REAL foo CALL foo() ! - wrong NAG f95 diagnoses: FOO has been declared as a function, not a subroutine ifort: The CALL statement is invoking a function subprogram as a subroutine. g95: Error: SUBROUTINE 'foo' at (1) cannot have REAL type -- Summary: External functions can be called as subroutine Product: gcc Version: 4.3.0 Status: UNCONFIRMED Keywords: accepts-invalid Severity: normal Priority: P3 Component: fortran AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: burnus at gcc dot gnu dot org http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33116
[Bug fortran/33117] New: Improve error message for generic interface with subroutines functions
Mixing functions and subroutines in a generic interface is invalid. gfortran detects this but the error message is difficult to understand: Error: FUNCTION at (1) does not belong in a generic subroutine interface NAG f95: Error: b.f90, line 10: BAR and SUB must both be SUBROUTINEs or both be FUNCTIONs for generic GEN g95: Error: Interfaces of generic 'gen' at (1) must all be FUNCTIONs or SUBROUTINEs ifort -stand f95: fortcom: Warning: b.f90, line 6: Standard Fortran requires that within a scoping unit, two procedures that have the same generic name shall both be subroutines or both be functions. [BAR] Example: module mod implicit none interface gen subroutine sub() end subroutine sub function bar() real :: bar end function bar end interface gen end module mod -- Summary: Improve error message for generic interface with subroutines functions Product: gcc Version: 4.3.0 Status: UNCONFIRMED Keywords: diagnostic Severity: normal Priority: P3 Component: fortran AssignedTo: unassigned at gcc dot gnu dot org ReportedBy: burnus at gcc dot gnu dot org http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33117
[Bug target/33115] -march=native is not supported under x86 darwin
--- Comment #1 from dje at gcc dot gnu dot org 2007-08-19 21:12 --- x86 darwin should be able to re-use the implementation from rs6000/driver-rs6000.c because the kernel interfaces to query the information are the same. -- dje at gcc dot gnu dot org changed: What|Removed |Added CC||dje at gcc dot gnu dot org http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33115
[Bug target/33115] -march=native is not supported under x86 darwin
--- Comment #2 from pinskia at gmail dot com 2007-08-19 21:39 --- Subject: Re: -march=native is not supported under x86 darwin On 19 Aug 2007 21:12:22 -, dje at gcc dot gnu dot org [EMAIL PROTECTED] wrote: x86 darwin should be able to re-use the implementation from rs6000/driver-rs6000.c because the kernel interfaces to query the information are the same. Well and i386/driver-i386.c uses inline-asm so it should able to use that also. EXTRA_GCC_OBJS =driver-i386.o darwin-driver.o We are able to compile driver-i386.c but just not use it. The problem is that CC1_SPEC does not use %(cc1_cpu). Thanks, Andrew Pinski -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33115
[Bug target/33115] -march=native is not supported under x86 darwin
--- Comment #3 from pinskia at gcc dot gnu dot org 2007-08-19 21:40 --- Mine as explained in comment #2 in how to fix this issue. -- pinskia at gcc dot gnu dot org changed: What|Removed |Added AssignedTo|unassigned at gcc dot gnu |pinskia at gcc dot gnu dot |dot org |org Status|UNCONFIRMED |ASSIGNED Ever Confirmed|0 |1 Last reconfirmed|-00-00 00:00:00 |2007-08-19 21:40:38 date|| http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33115
[Bug c/33111] Bad code generation with -O2 (ARM 7 architecture)
--- Comment #3 from gressau at optusnet dot com dot au 2007-08-19 22:22 --- Created an attachment (id=14077) -- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=14077action=view) This file does not compile correctly with -O2 -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33111
[Bug c/33111] Bad code generation with -O2 (ARM 7 architecture)
--- Comment #4 from gressau at optusnet dot com dot au 2007-08-19 22:24 --- Created an attachment (id=14078) -- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=14078action=view) This is the working unoptimized assembly of uip.c -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33111
[Bug c/33111] Bad code generation with -O2 (ARM 7 architecture)
--- Comment #5 from gressau at optusnet dot com dot au 2007-08-19 22:25 --- Created an attachment (id=14079) -- (http://gcc.gnu.org/bugzilla/attachment.cgi?id=14079action=view) This is the nonworking optimized assembly of uip.c -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33111
[Bug target/33115] -march=native is not supported under x86 darwin
--- Comment #4 from pinskia at gcc dot gnu dot org 2007-08-19 23:18 ---
Patch in testing:
Index: config/i386/darwin.h
===
--- config/i386/darwin.h(revision 127626)
+++ config/i386/darwin.h(working copy)
@@ -89,7 +89,7 @@ along with GCC; see the file COPYING3.
the kernel or some such. */
#undef CC1_SPEC
-#define CC1_SPEC %{!mkernel:%{!static:%{!mdynamic-no-pic:-fPIC}}} \
+#define CC1_SPEC %(cc1_cpu)%{!mkernel:%{!static:%{!mdynamic-no-pic:-fPIC}}} \
%{!mmacosx-version-min=*:-mmacosx-version-min=%(darwin_minversion)} \
%{g: %{!fno-eliminate-unused-debug-symbols: -feliminate-unused-debug-symbols
}}
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33115
[Bug tree-optimization/32772] [4.3 Regression] error: found real variable when subvariables should have appeared
--- Comment #8 from dberlin at gcc dot gnu dot org 2007-08-19 23:23 --- Subject: Bug 32772 Author: dberlin Date: Sun Aug 19 23:23:29 2007 New Revision: 127629 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=127629 Log: 2007-08-19 Daniel Berlin [EMAIL PROTECTED] Fix PR 32772 Fix PR 32716 Fix PR 32328 Fix PR 32303 * tree-flow.h (struct stmt_ann_d): Remove makes_clobbering_call. * tree-ssa-alias.c (init_transitive_clobber_worklist): Add on_worklist argument and avoid adding things to worklist multiple times. (add_to_worklist): Ditto. (mark_aliases_call_clobbered): Mark entire structure clobbered if single SFT is clobbered. (set_initial_properties): Ditto. (compute_call_clobbered): Update for changes to function arguments. (create_overlap_variables_for): Always create SFT for offset 0. (create_structure_vars): Handle PHI's, since we are in SSA form at this point. * tree-ssa-loop-ivopts.c (get_ref_tag): Don't return subvars. * tree-ssa-operands.c (access_can_touch_variable): Don't handle TARGET_MEM_REF. (add_vars_for_offset): Figure out aliases from access + points-to. (add_virtual_operand): Use add_vars-for_offset. (get_tmr_operands): Update for NMT changes, rewrite to be correct. (add_call_clobber_ops): Remove makes_clobbering_call set. (get_expr_operands): Always pass through the INDIRECT_REF reference. * tree-ssa-structalias.c (struct constraint_graph): Remove variables member. Add pe, pe_rep, pointer_label, loc_label, pointed_by, points_to, address_taken, pt_used, number_incoming. (FIRST_ADDR_NODE): Removed. (merge_graph_nodes): Remove broken code for the moment. (init_graph): New function. (build_pred_graph): Remove code to init_graph. Add location equivalence support. (struct scc_info): Rename roots to deleted. (scc_visit): Ditto. (init_scc_info): Ditto (init_topo_info): Use graph-size. (compute_topo_order): Ditto. (do_da_constraint): Removed. (do_sd_constraint): Remove calls to find(). set_union_with_increment should always get 0 as last arg here. (do_complex_constraint): Replace do_da_constraint with assert. Stop calling find. (struct equiv_class_label): New. (pointer_equiv_class_table): Ditto. (location_equiv_class_table): Ditto. (equiv_class_label_hash): Ditto. (equiv_class_label_eq): Ditto (equiv_class_lookup): Ditto. (equiv_class_ladd): Ditto. (pointer_equiv_class): Ditto. (location_equiv_class): Ditto. (condense_visit): Rename and rewrite from label_visit to do only SCC related stuff for HU. (label_visit): Do HU work for HU. (perform_var_substitution): Update to do HU and location equivalence. (free_var_substitution_info): Update to free HU and location equivalence structures. */ (find_equivalent_node): Update for pointer but not location equivalence. (unite_pointer_equivalences): New function. (move_complex_constraints): Rewrite to only do moving. (rewrite_constraints): Split out of move_complex_constraints. (solve_graph): Use graph-size. (process_constraint_1): Add from_call argument, use it. Split *a = b into two constraints. (process_constraint): Use new process_constraint_1. (get_constraint_for_component_ref): Handle bitmaxsize == -1 case. (get_constraint_for): Handle non-pointer integers properly. Remove code that used to handle structures. (handle_ptr_arith): Fix a few bugs in pointer arithmetic handling with unknown addends. (handle_rhs_call): New function. (find_func_aliases): Use handle_rhs_call. (set_uids_in_ptset): Add an assert. (set_used_smts): Fix bug in not considering unified vars. (compute_tbaa_pruning): Stop initing useless iteration_obstack. (compute_points_to_sets): Update for other function changes. (delete_points_to_sets): Ditto. (ipa_pta_execute): Ditto. (pass_ipa_pta): We need to update SSA after ipa_pta. Modified: trunk/gcc/ChangeLog trunk/gcc/tree-flow.h trunk/gcc/tree-ssa-alias.c trunk/gcc/tree-ssa-loop-ivopts.c trunk/gcc/tree-ssa-operands.c trunk/gcc/tree-ssa-structalias.c -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32772
[Bug tree-optimization/32303] [4.3 Regression] SPEC2006 447.dealII miscompiled at -O2
--- Comment #2 from dberlin at gcc dot gnu dot org 2007-08-19 23:23 --- Subject: Bug 32303 Author: dberlin Date: Sun Aug 19 23:23:29 2007 New Revision: 127629 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=127629 Log: 2007-08-19 Daniel Berlin [EMAIL PROTECTED] Fix PR 32772 Fix PR 32716 Fix PR 32328 Fix PR 32303 * tree-flow.h (struct stmt_ann_d): Remove makes_clobbering_call. * tree-ssa-alias.c (init_transitive_clobber_worklist): Add on_worklist argument and avoid adding things to worklist multiple times. (add_to_worklist): Ditto. (mark_aliases_call_clobbered): Mark entire structure clobbered if single SFT is clobbered. (set_initial_properties): Ditto. (compute_call_clobbered): Update for changes to function arguments. (create_overlap_variables_for): Always create SFT for offset 0. (create_structure_vars): Handle PHI's, since we are in SSA form at this point. * tree-ssa-loop-ivopts.c (get_ref_tag): Don't return subvars. * tree-ssa-operands.c (access_can_touch_variable): Don't handle TARGET_MEM_REF. (add_vars_for_offset): Figure out aliases from access + points-to. (add_virtual_operand): Use add_vars-for_offset. (get_tmr_operands): Update for NMT changes, rewrite to be correct. (add_call_clobber_ops): Remove makes_clobbering_call set. (get_expr_operands): Always pass through the INDIRECT_REF reference. * tree-ssa-structalias.c (struct constraint_graph): Remove variables member. Add pe, pe_rep, pointer_label, loc_label, pointed_by, points_to, address_taken, pt_used, number_incoming. (FIRST_ADDR_NODE): Removed. (merge_graph_nodes): Remove broken code for the moment. (init_graph): New function. (build_pred_graph): Remove code to init_graph. Add location equivalence support. (struct scc_info): Rename roots to deleted. (scc_visit): Ditto. (init_scc_info): Ditto (init_topo_info): Use graph-size. (compute_topo_order): Ditto. (do_da_constraint): Removed. (do_sd_constraint): Remove calls to find(). set_union_with_increment should always get 0 as last arg here. (do_complex_constraint): Replace do_da_constraint with assert. Stop calling find. (struct equiv_class_label): New. (pointer_equiv_class_table): Ditto. (location_equiv_class_table): Ditto. (equiv_class_label_hash): Ditto. (equiv_class_label_eq): Ditto (equiv_class_lookup): Ditto. (equiv_class_ladd): Ditto. (pointer_equiv_class): Ditto. (location_equiv_class): Ditto. (condense_visit): Rename and rewrite from label_visit to do only SCC related stuff for HU. (label_visit): Do HU work for HU. (perform_var_substitution): Update to do HU and location equivalence. (free_var_substitution_info): Update to free HU and location equivalence structures. */ (find_equivalent_node): Update for pointer but not location equivalence. (unite_pointer_equivalences): New function. (move_complex_constraints): Rewrite to only do moving. (rewrite_constraints): Split out of move_complex_constraints. (solve_graph): Use graph-size. (process_constraint_1): Add from_call argument, use it. Split *a = b into two constraints. (process_constraint): Use new process_constraint_1. (get_constraint_for_component_ref): Handle bitmaxsize == -1 case. (get_constraint_for): Handle non-pointer integers properly. Remove code that used to handle structures. (handle_ptr_arith): Fix a few bugs in pointer arithmetic handling with unknown addends. (handle_rhs_call): New function. (find_func_aliases): Use handle_rhs_call. (set_uids_in_ptset): Add an assert. (set_used_smts): Fix bug in not considering unified vars. (compute_tbaa_pruning): Stop initing useless iteration_obstack. (compute_points_to_sets): Update for other function changes. (delete_points_to_sets): Ditto. (ipa_pta_execute): Ditto. (pass_ipa_pta): We need to update SSA after ipa_pta. Modified: trunk/gcc/ChangeLog trunk/gcc/tree-flow.h trunk/gcc/tree-ssa-alias.c trunk/gcc/tree-ssa-loop-ivopts.c trunk/gcc/tree-ssa-operands.c trunk/gcc/tree-ssa-structalias.c -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32303
[Bug tree-optimization/32328] [4.2/4.3 Regression] -fstrict-aliasing causes skipped code
--- Comment #21 from dberlin at gcc dot gnu dot org 2007-08-19 23:23 --- Subject: Bug 32328 Author: dberlin Date: Sun Aug 19 23:23:29 2007 New Revision: 127629 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=127629 Log: 2007-08-19 Daniel Berlin [EMAIL PROTECTED] Fix PR 32772 Fix PR 32716 Fix PR 32328 Fix PR 32303 * tree-flow.h (struct stmt_ann_d): Remove makes_clobbering_call. * tree-ssa-alias.c (init_transitive_clobber_worklist): Add on_worklist argument and avoid adding things to worklist multiple times. (add_to_worklist): Ditto. (mark_aliases_call_clobbered): Mark entire structure clobbered if single SFT is clobbered. (set_initial_properties): Ditto. (compute_call_clobbered): Update for changes to function arguments. (create_overlap_variables_for): Always create SFT for offset 0. (create_structure_vars): Handle PHI's, since we are in SSA form at this point. * tree-ssa-loop-ivopts.c (get_ref_tag): Don't return subvars. * tree-ssa-operands.c (access_can_touch_variable): Don't handle TARGET_MEM_REF. (add_vars_for_offset): Figure out aliases from access + points-to. (add_virtual_operand): Use add_vars-for_offset. (get_tmr_operands): Update for NMT changes, rewrite to be correct. (add_call_clobber_ops): Remove makes_clobbering_call set. (get_expr_operands): Always pass through the INDIRECT_REF reference. * tree-ssa-structalias.c (struct constraint_graph): Remove variables member. Add pe, pe_rep, pointer_label, loc_label, pointed_by, points_to, address_taken, pt_used, number_incoming. (FIRST_ADDR_NODE): Removed. (merge_graph_nodes): Remove broken code for the moment. (init_graph): New function. (build_pred_graph): Remove code to init_graph. Add location equivalence support. (struct scc_info): Rename roots to deleted. (scc_visit): Ditto. (init_scc_info): Ditto (init_topo_info): Use graph-size. (compute_topo_order): Ditto. (do_da_constraint): Removed. (do_sd_constraint): Remove calls to find(). set_union_with_increment should always get 0 as last arg here. (do_complex_constraint): Replace do_da_constraint with assert. Stop calling find. (struct equiv_class_label): New. (pointer_equiv_class_table): Ditto. (location_equiv_class_table): Ditto. (equiv_class_label_hash): Ditto. (equiv_class_label_eq): Ditto (equiv_class_lookup): Ditto. (equiv_class_ladd): Ditto. (pointer_equiv_class): Ditto. (location_equiv_class): Ditto. (condense_visit): Rename and rewrite from label_visit to do only SCC related stuff for HU. (label_visit): Do HU work for HU. (perform_var_substitution): Update to do HU and location equivalence. (free_var_substitution_info): Update to free HU and location equivalence structures. */ (find_equivalent_node): Update for pointer but not location equivalence. (unite_pointer_equivalences): New function. (move_complex_constraints): Rewrite to only do moving. (rewrite_constraints): Split out of move_complex_constraints. (solve_graph): Use graph-size. (process_constraint_1): Add from_call argument, use it. Split *a = b into two constraints. (process_constraint): Use new process_constraint_1. (get_constraint_for_component_ref): Handle bitmaxsize == -1 case. (get_constraint_for): Handle non-pointer integers properly. Remove code that used to handle structures. (handle_ptr_arith): Fix a few bugs in pointer arithmetic handling with unknown addends. (handle_rhs_call): New function. (find_func_aliases): Use handle_rhs_call. (set_uids_in_ptset): Add an assert. (set_used_smts): Fix bug in not considering unified vars. (compute_tbaa_pruning): Stop initing useless iteration_obstack. (compute_points_to_sets): Update for other function changes. (delete_points_to_sets): Ditto. (ipa_pta_execute): Ditto. (pass_ipa_pta): We need to update SSA after ipa_pta. Modified: trunk/gcc/ChangeLog trunk/gcc/tree-flow.h trunk/gcc/tree-ssa-alias.c trunk/gcc/tree-ssa-loop-ivopts.c trunk/gcc/tree-ssa-operands.c trunk/gcc/tree-ssa-structalias.c -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32328
[Bug c++/32716] [4.2/4.3 Regression] Wrong code generation. Alias and C++ virtual bases problem.
--- Comment #8 from dberlin at gcc dot gnu dot org 2007-08-19 23:23 --- Subject: Bug 32716 Author: dberlin Date: Sun Aug 19 23:23:29 2007 New Revision: 127629 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=127629 Log: 2007-08-19 Daniel Berlin [EMAIL PROTECTED] Fix PR 32772 Fix PR 32716 Fix PR 32328 Fix PR 32303 * tree-flow.h (struct stmt_ann_d): Remove makes_clobbering_call. * tree-ssa-alias.c (init_transitive_clobber_worklist): Add on_worklist argument and avoid adding things to worklist multiple times. (add_to_worklist): Ditto. (mark_aliases_call_clobbered): Mark entire structure clobbered if single SFT is clobbered. (set_initial_properties): Ditto. (compute_call_clobbered): Update for changes to function arguments. (create_overlap_variables_for): Always create SFT for offset 0. (create_structure_vars): Handle PHI's, since we are in SSA form at this point. * tree-ssa-loop-ivopts.c (get_ref_tag): Don't return subvars. * tree-ssa-operands.c (access_can_touch_variable): Don't handle TARGET_MEM_REF. (add_vars_for_offset): Figure out aliases from access + points-to. (add_virtual_operand): Use add_vars-for_offset. (get_tmr_operands): Update for NMT changes, rewrite to be correct. (add_call_clobber_ops): Remove makes_clobbering_call set. (get_expr_operands): Always pass through the INDIRECT_REF reference. * tree-ssa-structalias.c (struct constraint_graph): Remove variables member. Add pe, pe_rep, pointer_label, loc_label, pointed_by, points_to, address_taken, pt_used, number_incoming. (FIRST_ADDR_NODE): Removed. (merge_graph_nodes): Remove broken code for the moment. (init_graph): New function. (build_pred_graph): Remove code to init_graph. Add location equivalence support. (struct scc_info): Rename roots to deleted. (scc_visit): Ditto. (init_scc_info): Ditto (init_topo_info): Use graph-size. (compute_topo_order): Ditto. (do_da_constraint): Removed. (do_sd_constraint): Remove calls to find(). set_union_with_increment should always get 0 as last arg here. (do_complex_constraint): Replace do_da_constraint with assert. Stop calling find. (struct equiv_class_label): New. (pointer_equiv_class_table): Ditto. (location_equiv_class_table): Ditto. (equiv_class_label_hash): Ditto. (equiv_class_label_eq): Ditto (equiv_class_lookup): Ditto. (equiv_class_ladd): Ditto. (pointer_equiv_class): Ditto. (location_equiv_class): Ditto. (condense_visit): Rename and rewrite from label_visit to do only SCC related stuff for HU. (label_visit): Do HU work for HU. (perform_var_substitution): Update to do HU and location equivalence. (free_var_substitution_info): Update to free HU and location equivalence structures. */ (find_equivalent_node): Update for pointer but not location equivalence. (unite_pointer_equivalences): New function. (move_complex_constraints): Rewrite to only do moving. (rewrite_constraints): Split out of move_complex_constraints. (solve_graph): Use graph-size. (process_constraint_1): Add from_call argument, use it. Split *a = b into two constraints. (process_constraint): Use new process_constraint_1. (get_constraint_for_component_ref): Handle bitmaxsize == -1 case. (get_constraint_for): Handle non-pointer integers properly. Remove code that used to handle structures. (handle_ptr_arith): Fix a few bugs in pointer arithmetic handling with unknown addends. (handle_rhs_call): New function. (find_func_aliases): Use handle_rhs_call. (set_uids_in_ptset): Add an assert. (set_used_smts): Fix bug in not considering unified vars. (compute_tbaa_pruning): Stop initing useless iteration_obstack. (compute_points_to_sets): Update for other function changes. (delete_points_to_sets): Ditto. (ipa_pta_execute): Ditto. (pass_ipa_pta): We need to update SSA after ipa_pta. Modified: trunk/gcc/ChangeLog trunk/gcc/tree-flow.h trunk/gcc/tree-ssa-alias.c trunk/gcc/tree-ssa-loop-ivopts.c trunk/gcc/tree-ssa-operands.c trunk/gcc/tree-ssa-structalias.c -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32716
[Bug middle-end/32940] REG_POINTER attribute on DECL_ARTIFICIAL pointers
--- Comment #7 from pinskia at gcc dot gnu dot org 2007-08-20 00:48 --- Subject: Bug 32940 Author: pinskia Date: Mon Aug 20 00:48:09 2007 New Revision: 127634 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=127634 Log: 2007-08-19 Andrew Pinski [EMAIL PROTECTED] PR middle-end/32940 * cfgexpand.c (expand_one_register_var): Mark pointer DECL_ARTIFICIAL as REG_POINTER also. * stmt.c (expand_decl): Likewise. Modified: trunk/gcc/ChangeLog trunk/gcc/cfgexpand.c trunk/gcc/stmt.c -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32940
[Bug middle-end/32940] REG_POINTER attribute on DECL_ARTIFICIAL pointers
--- Comment #8 from pinskia at gcc dot gnu dot org 2007-08-20 00:48 --- Fixed. -- pinskia at gcc dot gnu dot org changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution||FIXED Target Milestone|--- |4.3.0 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32940
[Bug testsuite/31884] priority_queue_dijkstra.cc operates on deallocated memory
--- Comment #3 from pinskia at gcc dot gnu dot org 2007-08-20 00:56 --- This test now works for me on i386-apple-darwin, it was failing before this patch. Fixed. -- pinskia at gcc dot gnu dot org changed: What|Removed |Added Status|UNCONFIRMED |RESOLVED Resolution||FIXED Target Milestone|--- |4.3.0 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=31884
[Bug middle-end/33092] [4.3 Regrsssion] Using -O1 -fno-tree-salias results in ICE
--- Comment #2 from pinskia at gcc dot gnu dot org 2007-08-20 01:46 --- Confirmed. -- pinskia at gcc dot gnu dot org changed: What|Removed |Added Status|UNCONFIRMED |NEW Ever Confirmed|0 |1 Last reconfirmed|-00-00 00:00:00 |2007-08-20 01:46:26 date|| http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33092
[Bug tree-optimization/32328] [4.2/4.3 Regression] -fstrict-aliasing causes skipped code
--- Comment #22 from dberlin at gcc dot gnu dot org 2007-08-20 01:54 --- All should be fixed -- dberlin at gcc dot gnu dot org changed: What|Removed |Added Status|NEW |RESOLVED Resolution||FIXED http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32328
[Bug tree-optimization/32303] [4.3 Regression] SPEC2006 447.dealII miscompiled at -O2
--- Comment #3 from dberlin at gcc dot gnu dot org 2007-08-20 01:54 --- All should be fixed -- dberlin at gcc dot gnu dot org changed: What|Removed |Added Status|NEW |RESOLVED Resolution||FIXED http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32303
[Bug tree-optimization/32772] [4.3 Regression] error: found real variable when subvariables should have appeared
--- Comment #9 from dberlin at gcc dot gnu dot org 2007-08-20 01:54 --- All should be fixed -- dberlin at gcc dot gnu dot org changed: What|Removed |Added Status|NEW |RESOLVED Resolution||FIXED http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32772
[Bug c++/32716] [4.2/4.3 Regression] Wrong code generation. Alias and C++ virtual bases problem.
--- Comment #9 from dberlin at gcc dot gnu dot org 2007-08-20 01:54 --- All should be fixed -- dberlin at gcc dot gnu dot org changed: What|Removed |Added Status|NEW |RESOLVED Resolution||FIXED http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32716
[Bug tree-optimization/32723] [4.2 Regression] memory hog in solve_graph
--- Comment #11 from dberlin at gcc dot gnu dot org 2007-08-20 01:56 --- Uh, it doesn't take 1 gig on either 4.2 or 4.3 -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=32723
[Bug c++/33118] New: [4.3 Regression] #'argument_pack_select' not supported by dump_expr#expression error
Testcase:
template typename _Tp
struct __add_const_lvalue_reference;
templatetypename... _Elements
class tuple
{
tuple(typename __add_const_lvalue_reference_Elements::type ...) ;
};
tupleint, int, int i;
- CUT -
Found while cutting down PR 33091. This is a regression because the error
message was ok in 4.2.0 and before.
On the trunk we get:
t.cc: In instantiation of 'tuple#'argument_pack_select' not supported by
dump_expr#expression error ':
t.cc:8: instantiated from here
t.cc:6: error: invalid use of incomplete type 'struct
__add_const_lvalue_referenceint'
--
Summary: [4.3 Regression] #'argument_pack_select' not supported
by dump_expr#expression error
Product: gcc
Version: 4.3.0
Status: UNCONFIRMED
Keywords: diagnostic
Severity: normal
Priority: P3
Component: c++
AssignedTo: unassigned at gcc dot gnu dot org
ReportedBy: pinskia at gcc dot gnu dot org
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33118
[Bug c++/33118] [4.3 Regression] #'argument_pack_select' not supported by dump_expr#expression error
-- pinskia at gcc dot gnu dot org changed: What|Removed |Added Target Milestone|--- |4.3.0 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33118
[Bug fortran/20896] [4.2 and 4.1 only] ambiguous interface not detected
--- Comment #10 from pault at gcc dot gnu dot org 2007-08-20 03:48 --- (In reply to comment #9) (In reply to comment #8) Paul, what is the status on this one? ping? Go back to the thread on it. I did not seem able to satisfy the reviewer as to my interpretation of the standard; since I could not see any other, I let it drop. I'll unassign myself. Paul -- pault at gcc dot gnu dot org changed: What|Removed |Added AssignedTo|pault at gcc dot gnu dot org|unassigned at gcc dot gnu ||dot org Status|ASSIGNED|NEW http://gcc.gnu.org/bugzilla/show_bug.cgi?id=20896
[Bug fortran/33116] External functions can be called as subroutine
--- Comment #1 from pault at gcc dot gnu dot org 2007-08-20 03:59 --- Tobias, EXTERNAL foo REAL foo CALL foo() ! - wrong END with 4.3.0 20070815, gfc gives: test.f90:1.14: EXTERNAL foo 1 test.f90:3.23: CALL foo() ! - wrong 2 Error: 'foo' at (1) has a type, which is not consistent with the CALL at (2) -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33116
[Bug target/33115] -march=native is not supported under x86 darwin
--- Comment #5 from pinskia at gcc dot gnu dot org 2007-08-20 05:16 --- Subject: Bug 33115 Author: pinskia Date: Mon Aug 20 05:16:10 2007 New Revision: 127635 URL: http://gcc.gnu.org/viewcvs?root=gccview=revrev=127635 Log: 2007-08-19 Andrew Pinski [EMAIL PROTECTED] PR target/33115 * config/i386/darwin.h (CC1_SPEC): Add %(cc1_cpu) in front. Modified: trunk/gcc/ChangeLog trunk/gcc/config/i386/darwin.h -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33115
[Bug target/33115] -march=native is not supported under x86 darwin
--- Comment #6 from pinskia at gcc dot gnu dot org 2007-08-20 05:30 --- Fixed (I added a space after %(cc1_cpu) for the final patch). -- pinskia at gcc dot gnu dot org changed: What|Removed |Added Status|ASSIGNED|RESOLVED Resolution||FIXED Target Milestone|--- |4.3.0 http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33115
[Bug tree-optimization/33113] Failing to represent the stride (with array) of a dataref when it is not a constant
--- Comment #2 from dorit at gcc dot gnu dot org 2007-08-20 05:55 --- Making us return symbolic stride would not be hard. The problem is that data dependence analysis would fail anyway, sometimes (not in this testcases) there won't be a need for dependence testing - e.g. a reduction computation where there are no stores, or initialization with a constant (i.e. a store and no loads), so there's already a value in doing this. since we cannot tell whether n is zero. can we do the data-dependence analysis conditioned on a maybe_zero (like the number-of-iterations analysis)? (by the way, I was told that ifort vectorizes this. I think we'd need loop reversal to vectorize the inner-loop though. on top of overcoming the unknown-stride issue in the DR and DDR analysis) -- http://gcc.gnu.org/bugzilla/show_bug.cgi?id=33113
