2.4.19-ck7
-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1
I've gone back and looked at ck7 (was a little while ago). -ck doesnt directly alter skbuff.c but may be responsible for calling it in an interrupt. I'm sorry I can't enlighten you as to why it's happening and offer a fix as I don't really know what the problem is.
The fact that it's happening now regularly and not previously is unusual if the kernel itself is responsible unless some pattern in your usage has changed. Perhaps seeing if the problem repeats on a vanilla or alternate kernel may be helpful (-ck is rather different from vanilla).
Con
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.2.0 (GNU/Linux)
iD8DBQE98EWvF6dfvkL3i1gRAnDbAJwN5YR8X1P/YR7XQLKAg+q1orm2SQCcDf+A i3iLphPRc5au7WsXBci+npA=
=JMES
-----END PGP SIGNATURE-----
Quoted 9 lines · Con Kolivas
I've gone back and looked at ck7 (was a little while ago). -ck doesnt directly alter
skbuff.cbut may be responsible for calling it in an interrupt. I'm sorry I can't enlighten you as to why it's happening and offer a fix as I don't really know what the problem is.The fact that it's happening now regularly and not previously is unusual if the kernel itself is responsible unless some pattern in your usage has changed. Perhaps seeing if the problem repeats on a vanilla or alternate kernel may be helpful (-ck is rather different from vanilla).
I suspect that the network load is getting high enough to travel new code paths, it's pushing ~200MB/s 24x7, and I'm sure we get some peaks as well. Just thought it might be useful, I will be going to something newer next week.
Signature · 3 lines
bill davidsen <[email protected]> CTO, TMR Associates, Inc Doing interesting things with little computers since 1979.