[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Old Index]
Re: Fixes for kern/40018 -- any chance of getting these pulled into the -current and 5.x trees?
Hello. I've tested on the following variants from that list:
BCM5706 or 5714
BCM5750 or BCM5721
In addition, this patch makes our driver drive these chips in the same
way that the FreeBSD, OpenBSD and Linux drivers do. the bug is that some
of the chip randomly switch modes and compute the hardware checksum on the
pseudo header even though we've instructed them not to. The other drivers
tell the chips to do the complete pseudo header calculation all the time.
I believe this patch is "correct" in so far as it now runs the chips in
a mode which is well tested and used by other *nix drivers -- in
particular -- drivers from Broadcom itself. From 2003 through today, our
driver runs the chip in a mode which appears to be completely unused, and,
from what I can tell, untested.
On Sep 21, 2:19pm, Thor Lancelot Simon wrote:
} Subject: Re: Fixes for kern/40018 -- any chance of getting these pulled in
} On Tue, Sep 21, 2010 at 10:05:52AM -0700, Brian Buhrow wrote:
} > Hello. I submitted a couple of patches for kern/40018, hardware
} > checksum failures on bge(4) controlled chips, which fix the problem. These
} Maybe they do, and maybe they don't. I would not trust these changes
} unless they have been verified on, at least:
} BCM5706 or 5714
} BCM5750 or BCM5721
} Have you tested all those hardware variants?
>-- End of excerpt from Thor Lancelot Simon
Main Index |
Thread Index |