[1165] in linux-net channel archive
8390 eth_copy_and_sum() improvements
daemon@ATHENA.MIT.EDU (Paul Gortmaker)
Wed Oct 4 12:13:58 1995
From: Paul Gortmaker <gpg109@rsphy1.anu.edu.au>
To: ftom@netcom.com (Tom May)
Date: Wed, 4 Oct 1995 16:15:26 +1000 (EST)
Cc: iialan@www.linux.org.uk (Alan Cox), linux-net@vger.rutgers.edu
In-Reply-To: <199507151933.MAA01173@netcom18.netcom.com> from "Tom May" at Jul 15, 95 12:33:15 pm
Hi Tom,
My eth_copy_and_sum() changes will be going into the 1.3.31 kernel for
the 8390 cards that can use it (wd, ultra, 3c503, e2100, ac3200). When
I originally did these patches about two months ago, you wrote:
> Before you benchmark, the csum_partial_copy() routine should be
> changed to use the inner loop from the original csum_partial_copyffs()
> function. I sent Alan a patch for this a while back. I suppose I
> should benchmark it to make sure it is worthwhile.
I was just wondering if this has been done yet? Was/is it worthwhile?
(I avoid assembly like the plague whenever possible ;-)
To linux-net people: please test these when they come out and make sure
I didn't break anything. The changes include the following:
- eth_io_copy_and_sum() for wd80x3, SMC-Ultra, 3c503, e2100, and ac3200
- ei_get_8390_hdr for all 8390 cards (to accomodate above)
- detect and accept new 3Com prefix for 3c503
- architecture indep. changes (i.e. readb(), memcpy_fromio(), HZ/100, etc.)
- tweak ne2k probe to help fools that use it as a module
- another ne2k clone added to the bad_clone_list
- eth_io versions aliased from macros in <asm-XXX/io.h> as Linus requested.
BTW, all 8390 cards should now work on the Alpha platforms as well (I hope).
Regards,
Paul.
