| To: | Christoph Hellwig <hch@xxxxxxxxxxxxx> |
|---|---|
| Subject: | Re: XFS causing stack overflow |
| From: | Andi Kleen <andi@xxxxxxxxxxxxxx> |
| Date: | Fri, 09 Dec 2011 11:53:36 -0800 |
| Cc: | "Ryan C. England" <ryan.england@xxxxxxxxxxxxx>, linux-mm@xxxxxxxxx, xfs@xxxxxxxxxxx |
| In-reply-to: | <20111209115513.GA19994__23079.9863501035$1323435203$gmane$org@xxxxxxxxxxxxx> (Christoph Hellwig's message of "Fri, 9 Dec 2011 06:55:13 -0500") |
| References: | <CAAnfqPAm559m-Bv8LkHARm7iBW5Kfs7NmjTFidmg-idhcOq4sQ@xxxxxxxxxxxxxx> <20111209115513.GA19994__23079.9863501035$1323435203$gmane$org@xxxxxxxxxxxxx> |
| User-agent: | Gnus/5.13 (Gnus v5.13) Emacs/23.3 (gnu/linux) |
Christoph Hellwig <hch@xxxxxxxxxxxxx> writes: > > You probably have only a third of the stack actually used by XFS, the > rest is from NFSD/writeback code and page reclaim. I don't think any > of this is easily fixable in a 2.6.32 codebase. Current mainline 3.2-rc > now has the I/O-less balance dirty pages which will basically split the > stack footprint in half, but it's an invasive change to the writeback > code that isn't easily backportable. An easy fix would be 16k stacks. Don't think they're that difficult to do, but would need a special binary. -Andi -- ak@xxxxxxxxxxxxxxx -- Speaking for myself only |
| <Prev in Thread] | Current Thread | [Next in Thread> |
|---|---|---|
| ||
| Previous by Date: | Re: [PATCH] xfstests 276: Ensure lost+found is not created with wrong link count, Eric Sandeen |
|---|---|
| Next by Date: | Re: XFS causing stack overflow, Dave Chinner |
| Previous by Thread: | Re: XFS causing stack overflow, Dave Chinner |
| Next by Thread: | [PATCH] xfstests: 277 new test to check freeze/unfreeze works well or not under heavy load., Masayoshi MIZUMA |
| Indexes: | [Date] [Thread] [Top] [All Lists] |