[Bug c++/71011] Wrong "may be uninitialized" warning
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71011 Pranith Kumar changed: What|Removed |Added CC||bobby.prani at gmail dot com --- Comment #1 from Pranith Kumar --- Created attachment 38443 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=38443&action=edit original source
[Bug c++/71011] New: Wrong "may be uninitialized" warning
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71011 Bug ID: 71011 Summary: Wrong "may be uninitialized" warning Product: gcc Version: 6.0 Status: UNCONFIRMED Severity: normal Priority: P3 Component: c++ Assignee: unassigned at gcc dot gnu.org Reporter: bobby.prani at gmail dot com Target Milestone: --- I am getting a warning that "data" may be uninitialized. But there is no such variable being used anywhere. src/allocate.cc: In member function ‘void allocate_c::run_a_cycle()’: src/allocate.cc:136:19: error: ‘data’ may be used uninitialized in this function [-Werror=maybe-uninitialized] else if (uop->m_uop_type == UOP_IADD || // integer register ~^~ cc1plus: all warnings being treated as errors Please find attached the source along with pre-processed file. Tested with g++ 4.7 and up and it is the same in all versions.
[Bug c++/71011] Wrong "may be uninitialized" warning
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71011 --- Comment #2 from Pranith Kumar --- Created attachment 38444 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=38444&action=edit pre-procssed source
[Bug c++/71011] Wrong "may be uninitialized" warning
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=71011 --- Comment #3 from Pranith Kumar --- Command which shows this error: $ g++ -o .opt_build/src/allocate.o -c -O3 -std=c++11 -funroll-loops -Werror -Wuninitialized -Wno-write-strings -DLONG_COUNTERS -DNO_MPI -DNO_DEBUG src/allocate.i -Wall -Wextra
[Bug c/79049] New: Unknown escape sequence not correctly pointed out
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=79049
Bug ID: 79049
Summary: Unknown escape sequence not correctly pointed out
Product: gcc
Version: 6.2.1
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c
Assignee: unassigned at gcc dot gnu.org
Reporter: bobby.prani at gmail dot com
Target Milestone: ---
GCC does not correctly point out the unknown escape sequence in the string
being printed.
#include
int main()
{
fprintf(stderr, "This statement has a \
backslash\ which is not correctly pointed out by gcc\n");
return 0;
}
Output from GCC:
$ gcc test.c -o test
test.c: In function ‘main’:
test.c:5:19: warning: unknown escape sequence: '\040'
fprintf(stderr, "This statement has a \
^~~
Output from Clang:
$ clang test.c -o test
test.c:6:29: warning: unknown escape sequence '\ ' [-Wunknown-escape-sequence]
backslash\ which is not correctly pointed out by gcc\n");
^~
[Bug c/66083] New: Linux kernel fails to boot with O3 optimization
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66083 Bug ID: 66083 Summary: Linux kernel fails to boot with O3 optimization Product: gcc Version: 4.9.2 Status: UNCONFIRMED Severity: normal Priority: P3 Component: c Assignee: unassigned at gcc dot gnu.org Reporter: bobby.prani at gmail dot com Target Milestone: --- Created attachment 35507 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=35507&action=edit Enable O3 optimization Linux kernel 4.1-rc1 fails to boot with O3 optimization. Patch to enable O3 and config file attached.
[Bug c/66083] Linux kernel fails to boot with O3 optimization
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66083 --- Comment #1 from Pranith Kumar --- Created attachment 35508 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=35508&action=edit config file
[Bug c/66083] Linux kernel fails to boot with O3 optimization
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=66083 --- Comment #3 from Pranith Kumar --- The kernel fails to boot. It works using 4.7/4.8 but 4.9 does not work. Not sure what you mean which file. How do I pin point that?
[Bug rtl-optimization/63375] [4.8/4.9/5 Regression] reordering of reads across fences
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63375 --- Comment #11 from Pranith Kumar --- Is there any work around for this in the mean time? Thanks!
[Bug rtl-optimization/63375] [4.8/4.9/5 Regression] reordering of reads across fences
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63375
--- Comment #13 from Pranith Kumar ---
The main concern here is moving the read past the fence instruction
irrespective of volatile semantics. The fence instruction guarantees that
accesses before the fence will complete before the accesses coming after the
fence. Consider the following case:
#include
typedef struct {
int counter;
} atomic_t;
static inline int atomic_read(atomic_t *v)
{
return (*(volatile int *)&(v)->counter);
}
static inline void atomic_write(atomic_t *v, int val)
{
v->counter = val;
}
#define smp_mb() asm volatile ("mfence":::"memory")
atomic_t val2 = {1};
int main()
{
atomic_t val1 = {1};
int p, q;
smp_mb();
p = atomic_read(&val1);
smp_mb();
atomic_write(&val2, 2);
smp_mb();
q = atomic_read(&val2);
printf("%d %d\n", p, q);
return 0;
}
Here, because of the bug the read from val1 is being generated after the write
to val2 breaking the semantics of memory fences. What am I missing?
[Bug rtl-optimization/63375] [4.8/4.9/5 Regression] reordering of reads across fences
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63375
--- Comment #16 from Pranith Kumar ---
This is not a good use case(it is a bit twisted, invalid too maybe?), but when
I try to read the stack without aliasing, the volatile write which I performed
is not visible.
#include
#include
#include
typedef struct {
int counter;
} atomic_t;
static inline int atomic_read(atomic_t *v)
{
return (*(volatile int *)&(v)->counter);
}
static inline void atomic_write(atomic_t *v, int val)
{
v->counter = val;
}
#define smp_mb() asm volatile ("mfence":::"memory")
int i = 0;
int main(int argc, char *argv[])
{
atomic_t val1 = {2};
atomic_t val2 = {5};
int p = 0;
printf("%p %p\n", &val1, &val2); // try commenting this
smp_mb();
p = atomic_read(&val1);
smp_mb();
atomic_write(&val2, 9);
smp_mb();
printf("%d\n", *(int *)(&val1+4));
return 0;
}
In the above program, if I comment the first printf() there is no alias for
val2 and the atomic_write() is optimized out and I see 0 printed out in the
second printf(). But without commenting the first printf() the write is issued
because of which I see 9 printed in the second printf.
[Bug rtl-optimization/63375] New: reordering of reads across fences
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63375
Bug ID: 63375
Summary: reordering of reads across fences
Product: gcc
Version: 4.9.1
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: rtl-optimization
Assignee: unassigned at gcc dot gnu.org
Reporter: bobby.prani at gmail dot com
A read within the region of code enclosed by barriers is being moved out of the
regions. Test case follows:
#include
typedef struct {
int counter;
} atomic_t;
static inline int atomic_read(atomic_t *v)
{
return (*(volatile int *)&(v)->counter);
}
#define smp_mb() asm volatile ("mfence":::"memory")
int main()
{
atomic_t val = {1};
int p;
smp_mb();
p = atomic_read(&val);
smp_mb();
printf("%d\n", p);
return 0;
}
Generated assembly:
15:mfence
18:mfence
1b:mov$0x1,%esi
20:mov$0x0,%rdi
27:xor%eax,%eax
29:callq 2e
[Bug rtl-optimization/63375] reordering of reads across fences
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63375 Pranith Kumar changed: What|Removed |Added CC||bobby.prani at gmail dot com --- Comment #1 from Pranith Kumar --- Created attachment 33576 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=33576&action=edit test case $ gcc -S -O2 test.c
[Bug c++/63444] New: Compilation consumes 2.5G memory
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63444 Bug ID: 63444 Summary: Compilation consumes 2.5G memory Product: gcc Version: 4.9.1 Status: UNCONFIRMED Severity: normal Priority: P3 Component: c++ Assignee: unassigned at gcc dot gnu.org Reporter: bobby.prani at gmail dot com Created attachment 33641 --> https://gcc.gnu.org/bugzilla/attachment.cgi?id=33641&action=edit pre-processed file The compilation of the attached pre-processed file takes 2.5G memory. Not exactly a regression since both 4.8 and 4.9 do the same. But may be an interesting test case for optimizations. $ g++ -std=c++11 -c -fPIC -DNDEBUG -O3 -fno-strict-aliasing -D_REENTRANT -D_POSIX_PTHREAD_SEMANTICS -D_FILE_OFFSET_BITS=64 test.i -o test.o
[Bug c++/63444] Compilation consumes 2.5G memory
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=63444 --- Comment #1 from Pranith Kumar --- Just FYI, clang compiles the same file using 1G memory.
[Bug c++/68821] New: g++ does not warn/error when virtual function return type is different than the one it overrides
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=68821
Bug ID: 68821
Summary: g++ does not warn/error when virtual function return
type is different than the one it overrides
Product: gcc
Version: unknown
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: c++
Assignee: unassigned at gcc dot gnu.org
Reporter: bobby.prani at gmail dot com
Target Milestone: ---
Consider:
typedef void (*obj_t)(void);
class parent {
public:
struct obj_base {
virtual ~obj_base() {}
virtual void operator()(void)=0;
};
template struct obj : public obj_base {
typedef void (T::*obj_t)(void);
T* p; obj_t f;
obj(T* p, obj_t f) : p(p), f(f) {}
int *operator()(void) {
return 0;
}
};
};
The overridden function has a different return type in the two cases. g++ does
not warn/error out about this.
