Re: [PATCH v1 1/7] elf-ops: bail out if we have no function symbols

["Philippe Mathieu-Daudé" <philmd@redhat.com>]

On 3/27/20 11:53 AM, Philippe Mathieu-Daudé wrote:
> On 3/27/20 10:49 AM, Alex Bennée wrote:
>> It's perfectly possible to have no function symbols in your elf file
>> and if we do the undefined behaviour sanitizer rightly complains about
>> us passing NULL to qsort. Check nsyms before we go ahead.
>>
>> Signed-off-by: Alex Bennée <alex.bennee@linaro.org>
>> ---
>>   include/hw/elf_ops.h | 7 ++++++-
>>   1 file changed, 6 insertions(+), 1 deletion(-)
>>
>> diff --git a/include/hw/elf_ops.h b/include/hw/elf_ops.h
>> index a1411bfcab6..b5d4074d1e3 100644
>> --- a/include/hw/elf_ops.h
>> +++ b/include/hw/elf_ops.h
>> @@ -170,8 +170,13 @@ static int glue(load_symbols, SZ)(struct elfhdr 
>> *ehdr, int fd, int must_swab,
>>           }
>>           i++;
>>       }
>> -    syms = g_realloc(syms, nsyms * sizeof(*syms));

Something was bugging me why looking at this line, now I remembered: 
another patch from 2 years ago :)

https://www.mail-archive.com/qemu-devel@nongnu.org/msg528713.html

Is this the same emitted warning? It seems.

   $ qemu-system-xtensa -M kc705 -m 128M -semihosting -nographic 
-monitor null -kernel Image.elf
   include/hw/elf_ops.h:179:5: runtime error: null pointer passed as 
argument 1, which is declared to never be null

If so, can you add it to the commit description?

Thanks!

>> +    /* check we have symbols left */
>> +    if (nsyms == 0) {
>> +        goto fail;
>> +    }
>> +
>> +    syms = g_realloc(syms, nsyms * sizeof(*syms));
>>       qsort(syms, nsyms, sizeof(*syms), glue(symcmp, SZ));
>>       for (i = 0; i < nsyms - 1; i++) {
>>           if (syms[i].st_size == 0) {
>>
> 
> Reviewed-by: Philippe Mathieu-Daudé <philmd@redhat.com>