public inbox for gcc-bugs@sourceware.org
help / color / mirror / Atom feed
* [Bug middle-end/94580] New: missing warning accessing an interior flexible array member
@ 2020-04-13 17:46 msebor at gcc dot gnu.org
2020-04-13 17:48 ` [Bug middle-end/94580] " msebor at gcc dot gnu.org
0 siblings, 1 reply; 2+ messages in thread
From: msebor at gcc dot gnu.org @ 2020-04-13 17:46 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94580
Bug ID: 94580
Summary: missing warning accessing an interior flexible array
member
Product: gcc
Version: 10.0
Status: UNCONFIRMED
Severity: normal
Priority: P3
Component: middle-end
Assignee: unassigned at gcc dot gnu.org
Reporter: msebor at gcc dot gnu.org
Target Milestone: ---
In the following (invalid) test case -Warray-bounds only diagnoses one of the
four out-of-bounds references, although it eliminates the seemingly redundant
stores on the (clearly incorrect in this case) assumption that they don't alias
with the reads. It's fine to eliminate the stores (or take whatever other
reasonable action) as long as the out-of-bounds accesses are diagnosed.
GCC warns about declarations of interior flexible array members, but only with
-Wpedantic, so this kind of bugs can easily go undetected.
$ cat t.c && gcc -O2 -S -Wall -Wextra -fdump-tree-vrp=/dev/stdout t.c
struct A { int n, a[]; };
struct B { struct A a; int x; };
struct B b;
int f0 (void)
{
b.a.a[1] = 1; // missing -Warray-bounds (store eliminated)
int t = b.x;
b.a.a[1] = 0; // -Warray-bounds (good)
return b.x - t;
}
int f1 (struct B *p)
{
p->a.a[1] = 1; // missing -Warray-bounds (store eliminated)
int t = p->x;
p->a.a[1] = 0; // missing -Warray-bounds
return p->x - t;
}
;; Function f0 (f0, funcdef_no=0, decl_uid=1937, cgraph_uid=1, symbol_order=1)
;; 1 loops found
;;
;; Loop 0
;; header 0, latch 1
;; depth 0, outer -1
;; nodes: 0 1 2
;; 2 succs { 1 }
Value ranges after VRP:
t.c: In function ‘f0’:
t.c:10:8: warning: array subscript 1 is above array bounds of ‘int[0]’
[-Warray-bounds]
10 | b.a.a[1] = 0; // -Warray-bounds (good)
| ~~~~~^~~
t.c:1:19: note: while referencing ‘a’
1 | struct A { int n, a[]; };
| ^
f0 ()
{
<bb 2> [local count: 1073741824]:
b.a.a[1] = 0;
return 0;
}
;; Function f0 (f0, funcdef_no=0, decl_uid=1937, cgraph_uid=1, symbol_order=1)
;; 1 loops found
;;
;; Loop 0
;; header 0, latch 1
;; depth 0, outer -1
;; nodes: 0 1 2
;; 2 succs { 1 }
Value ranges after VRP:
f0 ()
{
<bb 2> [local count: 1073741824]:
b.a.a[1] = 0;
return 0;
}
;; Function f1 (f1, funcdef_no=1, decl_uid=1941, cgraph_uid=2, symbol_order=2)
;; 1 loops found
;;
;; Loop 0
;; header 0, latch 1
;; depth 0, outer -1
;; nodes: 0 1 2
;; 2 succs { 1 }
Value ranges after VRP:
f1 (struct B * p)
{
<bb 2> [local count: 1073741824]:
p_2(D)->a.a[1] = 0;
return 0;
}
;; Function f1 (f1, funcdef_no=1, decl_uid=1941, cgraph_uid=2, symbol_order=2)
;; 1 loops found
;;
;; Loop 0
;; header 0, latch 1
;; depth 0, outer -1
;; nodes: 0 1 2
;; 2 succs { 1 }
Value ranges after VRP:
f1 (struct B * p)
{
<bb 2> [local count: 1073741824]:
p_2(D)->a.a[1] = 0;
return 0;
}
^ permalink raw reply [flat|nested] 2+ messages in thread
* [Bug middle-end/94580] missing warning accessing an interior flexible array member
2020-04-13 17:46 [Bug middle-end/94580] New: missing warning accessing an interior flexible array member msebor at gcc dot gnu.org
@ 2020-04-13 17:48 ` msebor at gcc dot gnu.org
0 siblings, 0 replies; 2+ messages in thread
From: msebor at gcc dot gnu.org @ 2020-04-13 17:48 UTC (permalink / raw)
To: gcc-bugs
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=94580
Martin Sebor <msebor at gcc dot gnu.org> changed:
What |Removed |Added
----------------------------------------------------------------------------
Keywords| |diagnostic
Blocks| |56456
--- Comment #1 from Martin Sebor <msebor at gcc dot gnu.org> ---
GCC does a little better with zero-length arrays:
$ cat t.c && gcc -O2 -S -Wall -Wextra -o/dev/stdout t.c
struct A { int n, a[0]; };
struct B { struct A a; int x; };
struct B b;
int f0 (void)
{
b.a.a[1] = 1;
int t = b.x;
b.a.a[1] = 0;
return b.x - t;
}
int f1 (struct B *p)
{
p->a.a[1] = 1;
int t = p->x;
p->a.a[1] = 0;
return p->x - t;
}
.file "t.c"
.text
t.c: In function ‘f0’:
t.c:10:8: warning: array subscript 1 is above array bounds of ‘int[0]’
[-Warray-bounds]
10 | b.a.a[1] = 0;
| ~~~~~^~~
t.c:1:19: note: while referencing ‘a’
1 | struct A { int n, a[0]; };
| ^
.p2align 4
.globl f0
.type f0, @function
f0:
.LFB0:
.cfi_startproc
movl $0, b+8(%rip)
xorl %eax, %eax
ret
.cfi_endproc
.LFE0:
.size f0, .-f0
t.c: In function ‘f1’:
t.c:18:9: warning: array subscript 1 is outside the bounds of an interior
zero-length array ‘int[0]’ [-Wzero-length-bounds]
18 | p->a.a[1] = 0;
| ~~~~~~^~~
t.c:1:19: note: while referencing ‘a’
1 | struct A { int n, a[0]; };
| ^
.p2align 4
.globl f1
.type f1, @function
f1:
.LFB1:
.cfi_startproc
movl $0, 8(%rdi)
xorl %eax, %eax
ret
.cfi_endproc
.LFE1:
.size f1, .-f1
.globl b
.bss
.align 8
.type b, @object
.size b, 8
b:
.zero 8
.ident "GCC: (GNU) 10.0.1 20200413 (experimental)"
.section .note.GNU-stack,"",@progbits
Referenced Bugs:
https://gcc.gnu.org/bugzilla/show_bug.cgi?id=56456
[Bug 56456] [meta-bug] bogus/missing -Warray-bounds
^ permalink raw reply [flat|nested] 2+ messages in thread
end of thread, other threads:[~2020-04-13 17:48 UTC | newest]
Thread overview: 2+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2020-04-13 17:46 [Bug middle-end/94580] New: missing warning accessing an interior flexible array member msebor at gcc dot gnu.org
2020-04-13 17:48 ` [Bug middle-end/94580] " msebor at gcc dot gnu.org
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).