cxgb - driver fixes.

Jeff,

I'm submitting a patch series for inclusion in 2.6.24 for the cxgb driver. The patches are built against Linus'git tree.

Here is a brief description:
- Ensure that GSO skbs have enough headroom before encapsulating them,
- Fix a crash in NAPI mode,
- Fix statistics accounting and report.

Quoted 9 lines · Divy Le Ray

Jeff,

I'm submitting a patch series for inclusion in 2.6.24 for the cxgb driver. The patches are built against Linus'git tree.

Here is a brief description:
- Ensure that GSO skbs have enough headroom before encapsulating them,
- Fix a crash in NAPI mode,
- Fix statistics accounting and report.

We ran pktgen overnight on 2.6.23 with patch 1 and 3 applied (patch 2 not needed on .23 it seems) and it was stable at about 1.5Gbps bi-directional using 1500 MTU sized frames.

We'll run some more tests with user-space TCP & UDP today, but it looks good so far.

Perhaps these patches should be considered for .23 stable as well?

Thanks,
Ben

Quoted 23 lines · Ben Greear

Quoted 10 lines · Divy Le Ray

Jeff,

I'm submitting a patch series for inclusion in 2.6.24 for the cxgb driver.
The patches are built against Linus'git tree.

Here is a brief description:
- Ensure that GSO skbs have enough headroom before encapsulating them,
- Fix a crash in NAPI mode,
- Fix statistics accounting and report.

We ran pktgen overnight on 2.6.23 with patch 1 and 3 applied (patch 2 not needed on .23 it seems) and it was stable at about 1.5Gbps bi-directional
using 1500 MTU sized frames.

We'll run some more tests with user-space TCP & UDP today, but it looks good
so far.

Perhaps these patches should be considered for .23 stable as well?

Quoted 3 lines

Thanks,
Ben

Yes, patches 1 and 3 could be considered for .22 and .23-stable, along with the previous fix:
http://git.kernel.org/?p=linux/kernel/git/torvalds/linux-2.6.git;a=commitdiff;h=7de6af0f23b25df8da9719ecae1916b669d0b03d