Re: Simple fix for swap_out (the 1st August patch)

Bill Hawes (
Fri, 01 Aug 1997 12:41:26 -0400

Dr. Werner Fink wrote:
> NOTE: The appended patch is against pre-patch-2.0.31-2.

I can't comment on the swap code changes (haven't studied that yet), but
I would like to point out what I believe is a mistake in
In 2.0.29 and 2.0.30 buffer.c the condition for waking up bdflush looks

if (!(grow_buffers(GFP_ATOMIC,size)))
needed -= PAGE_SIZE;
goto repeat;

This was changed to something like:
if ((grow_buffers(GFP_ATOMIC,size)))
needed -= PAGE_SIZE;
goto repeat;

On the surface the original code looks erroneous, but intentional or
not, the original code has the effect of reducing the memory goal _even
if no memory was allocated_. This means that tasks trapped in
"refill_freelist hell" have a chance of escaping under 2.0.29 or
unpatched 2.0.30, but may be forever trapped under the new code.

If you've looked at the changes I've made to 2.1.47 fs/buffer.c, it's
important to get tasks out of refill_freelist even if they haven't
reached their full quota of buffers. Otherwise, you can reach a
deadlock state where the tasks wanting buffers have all fallen into
refill_freelist, and none can make their quota.

I would much prefer to see the code written as something like:
if (free_list[BUFSIZE_INDEX(size)] != NULL)
needed -= PAGE_SIZE;
goto repeat;

i.e a direct test of success, but at the very least would recommend that
the original code be used.