Re: inode highmem imbalance fix [Re: Bug with shared memory.]

[William Lee Irwin III]

On Fri, May 24, 2002 at 09:33:41AM +0200, Andrea Arcangeli wrote:
> Here it is, you should apply it together with vm-35 that you need too
> for the bh/highmem balance (or on top of 2.4.19pre8aa3). I tested it
> slightly on uml and it didn't broke so far, so be careful because it's not
> very well tested yet. On the lines of what Alexey suggested originally,
> if goal isn't reached, in a second pass we shrink the cache too, but
> only if the cache is the only reason for the "pinning" beahiour of the
> inode. If for example there are dirty blocks of metadata or of data
> belonging to the inode we wakeup_bdflush instead and we never shrink the
> cache in such case. If the inode itself is dirty as well we let the two
> passes fail so we will schedule the work for keventd. This logic should
> ensure we never fall into shrinking the cache for no good reason and
> that we free the cache only for the inodes that we actually go ahead and
> free. (basically only dirty pages set with SetPageDirty aren't trapped
> by the logic before calling the invalidate, like ramfs, but that's
> expected of course, those pages cannot be damaged by the non destructive
> invalidate anyways)
> Comments?

I haven't had the chance to give this a test run yet, but it looks very
promising. I have a slight concern about the hold time of the inode_lock
because prune_icache() already generates some amount of contention,
but what you've presented appears to be necessary to prevent lethal
cache bloat, and so that concern is secondary at most. I'll give it a
test run tomorrow if no one else on-site gets to it first, though with
the proviso that I've not been involved in workloads triggering this
specific KVA exhaustion issue, so what testing I can do is limited.

Thanks,
Bill