On 25/11/16 18:50, Paul Eggert wrote:
> Pádraig Brady wrote:
>> for UBSAN we should probably build with
>> _STRING_ARCH_unaligned defined globally
>> to avoid warning for the cases we already handle.
>
> Yes. Translating this for non-experts: the problem here is a bug in the
> bug-finding procedu
Pádraig Brady wrote:
for UBSAN we should probably build with
_STRING_ARCH_unaligned defined globally
to avoid warning for the cases we already handle.
Yes. Translating this for non-experts: the problem here is a bug in the
bug-finding procedure, not a bug in GNU coreutils or in Gnulib.
Recen
On 25/11/16 06:18, Marcel Böhme wrote:
> Dear all,
>
> The following execution is flagged by UBSAN as undefined behaviour:
>
> $ echo 0 > a; printf "%0.s0" {1..58} >> a
> $ ./sort -R a
>
> UBSAN says:
> ../lib/md5.c:371:7: runtime error: load of misaligned address 0x7ffdfd45a10d
> for type 'con
Dear all,
The following execution is flagged by UBSAN as undefined behaviour:
$ echo 0 > a; printf "%0.s0" {1..58} >> a
$ ./sort -R a
UBSAN says:
../lib/md5.c:371:7: runtime error: load of misaligned address 0x7ffdfd45a10d
for type 'const uint32_t', which requires 4 byte alignment
So, the roo
