Re: CFT: re(4)

On Sat, Jun 02, 2007 at 05:10:09PM +0900, To freebsd-current_at_FreeBSD.org wrote:
 > On Tue, May 29, 2007 at 09:18:37PM +0900, To freebsd-current_at_FreeBSD.org wrote:
 >  > 
 >  > Dear all,
 >  > 
 >  > I've committed a fix for bus_dma(9) bug which resulted in poor Tx
 >  > performance on TSO enabled re(4) driver. With the fix and revised
 >  > re(4) I got more sane performance on re(4). Because there are too many
 >  > hardwares that rely on re(4) I'd like to hear any success or failure
 >  > reports before revised re(4) hits the tree.
 >  > For PCIe hardware users it would be great if you can submit
 >  > performance numbers for stock re(4) and revised one. The revised
 >  > re(4) can be found at the following URL.
 >  > http://people.freebsd.org/~yongari/re/re.HEAD.patch
 >  > 
 >  > Note, you need latest kernel to get correct performance numbers.
 >  > 
 > 
 > I've fixed a bug which resulted in checksum offload bug and update the
 > patch. It should have no ression.
 > We're very close to code freeze and have too many consumers of re(4).
 > Without users success report it would be impossible to commit the patch
 > before branching 7. Since I don't have 8139C+ based ones, I'm also
 > interested in how it works on 8139C+ hardwares.
 > 

I received few feedbacks on overhauled re(4).
Without this patch TSO would be unstable and you would get
"can't map defragmented mbuf" messages on console under heavy load.
If I couldn't get more feedbacks I'll disable TSO support before code
freeze.

-- 
Regards,
Pyun YongHyeon