https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108189
--- Comment #14 from Daniel Xu ---
Yeah, I saw that. But that only makes the build not
fail right? For bpftrace we are trying hard to be
warning-free.
FWIW I tried adding `-std=gnu2x` and it didn't help.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108189
--- Comment #13 from Jose E. Marchesi ---
In bpf-next we are passing -Wno-error for the particular BPF selftests that use
this construct:
progs/btf_dump_test_case_bitfields.c-CFLAGS := -Wno-error
progs/btf_dump_test_case_namespacing.c-CFLAGS :=
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108189
Daniel Xu changed:
What|Removed |Added
CC||dxu at dxuuu dot xyz
--- Comment #12 from D
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108189
--- Comment #11 from James Hilliard ---
(In reply to Andrew Pinski from comment #1)
> The error is correct at least for C before C23.
Can you clarify where exactly in the C23 specification that this will be
allowed?
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108189
--- Comment #10 from Andrew Pinski ---
(In reply to James Hilliard from comment #8)
> Yeah, clang with -std=gnu17 -pedantic doesn't even complain at all from the
> looks of it. Does it make sense to allow disabling this specific warning in
> GC
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108189
--- Comment #9 from James Hilliard ---
(In reply to Andrew Pinski from comment #5)
> Also Clang does not implement this warning at all.
>
> It is a bug in bpf-next really.
What would be the correct way to fix these in bpf-next?
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108189
--- Comment #8 from James Hilliard ---
(In reply to Andrew Pinski from comment #7)
> Yes the warning is really still correct even if a closer testcase would be:
> ```
> int f(struct {int t;} *b)
> {
> return b->t;
> }
>
> int f1(void *a)
> {
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108189
--- Comment #7 from Andrew Pinski ---
Yes the warning is really still correct even if a closer testcase would be:
```
int f(struct {int t;} *b)
{
return b->t;
}
int f1(void *a)
{
return f(a);
}
```
I am actually shocked clang didn't impleme
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108189
Andrew Pinski changed:
What|Removed |Added
Status|UNCONFIRMED |RESOLVED
Resolution|---
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108189
--- Comment #5 from Andrew Pinski ---
Also Clang does not implement this warning at all.
It is a bug in bpf-next really.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108189
--- Comment #4 from Andrew Pinski ---
Also Clang does not implement this warning at all.
It is a bug in bpf-next really.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108189
--- Comment #3 from Andrew Pinski ---
GCC does not implement all of C2X yet.
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108189
--- Comment #2 from James Hilliard ---
(In reply to Andrew Pinski from comment #1)
> The error is correct at least for C before C23.
Hmm, seeing it with -std=gnu2x passed still:
/home/buildroot/opt/cross/bin/bpf-gcc -g -Werror -D__TARGET_ARCH_
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=108189
--- Comment #1 from Andrew Pinski ---
The error is correct at least for C before C23.
14 matches
Mail list logo
