On Friday 13 January 2006 21:19, Henrik Nordstrom wrote:
> On Fri, 13 Jan 2006, H wrote:
> >> shm and msg sizes needs to be well within the requirements for diskd
> >> spelled out in the FAQ.
> >
> > nothing new but this thread is about the problem stated in the subject
>
> Right.. the answer is almost..
>
> The correct answer is to look into the q1/q2 parameers of your diskd
> cache_dir line. These MUST NOT be too large as there is a hardcoded
> internal limit of 96 concurrent requests per diskd.
>
i don't know since the assertion failure ocurres independent of q1/q2 values
after some time this function returns a bad buf value but why?
for (i = 0; i < SHMBUFS; i++) {
if (CBIT_TEST(diskdinfo->shm.inuse_map, i))
continue;
CBIT_SET(diskdinfo->shm.inuse_map, i);
*shm_offset = i * SHMBUF_BLKSZ;
buf = diskdinfo->shm.buf + (*shm_offset);
break;
}
--
H.
A mensagem foi scaneada pelo sistema de e-mail e pode ser considerada segura.
Service fornecido pelo Datacenter Matik https://datacenter.matik.com.br
