At Tue, 28 Sep 2004 18:33:57 +0300,
Ognyan Kulev wrote:
>
> Neal H. Walfield wrote:
> > This looks correct to me. Normally, I would say that this sort of
> > micro-optimization does not matter as we don't care how slow the error
> > path is, however, the other jump to allow_term_out (on the fast
Neal H. Walfield wrote:
This looks correct to me. Normally, I would say that this sort of
micro-optimization does not matter as we don't care how slow the error
path is, however, the other jump to allow_term_out (on the fast path)
also has a gratuitous unlock relock sequence:
/* Let someone else
At Tue, 28 Sep 2004 14:07:52 +0300,
Ognyan Kulev wrote:
>
> 2004-09-28 Ognyan Kulev <[EMAIL PROTECTED]>
>
> * data-request.c (_pager_seqnos_memory_object_data_request):
> When pager state is not NORMAL, shorten exit path.
> When _pager_pagemap_resize fails, add call to
>
2004-09-28 Ognyan Kulev <[EMAIL PROTECTED]>
* data-request.c (_pager_seqnos_memory_object_data_request):
When pager state is not NORMAL, shorten exit path.
When _pager_pagemap_resize fails, add call to
_pager_allow_termination.
diff -urpN --exclude='*~' --exclude=b
