Subject: Issue 110 of the NetBSD CVS Digest is out
To: None <email@example.com>
From: Mark Kirby <firstname.lastname@example.org>
Date: 09/30/2007 11:52:55
Joerg Sonnenberger and Jared D. McNeil have been working away at acpi =20=
support for AMD64. They have just put out a request for testers so if =20=
you are able please test their work.
Apply a patch from openssl CVS to fix a remaining off-by-one error in =20=
an older security fix, see link to the vulnerability . [Matthias =20
x86 changes for pcc and LKMs. [Andew Doran]
* Replace most inline assembly with proper functions. As a side =20
effect this reduces the size of amd64 GENERIC by about 120kB, and =20
i386 by a smaller amount. Nearly all of the inlines did something =20
slow, or something that does not need to be fast.
* Make curcpu() and curlwp functions proper, unless GNUC && _KERNEL. =20
In that case make them inlines. Makes curlwp LKM and preemption safe.
* Make bus_space and bus_dma more LKM friendly.
* Share a few more files between the ports.
* Other minor changes.
Nfe(4) has become more stable. [Quentin Garnier]
* If the chip doesn=92t support jumbo frames, don=92t use =20
bus_dmamap_load_mbuf which expects a properly filled mbuf chain, but =20
bus_dmamap_load for the mbuf storage space instead.
* If the chip supports jumbo frames
1 keep track of which RX descriptor uses which jumbo mbuf buffer, =20
so that we can rewrite the physaddr field of the descriptor later, as =20=
it might be partially overwritten by the hw
2 when we=92re out of jumbo mbufs, and if the packet is small =20
enough, copy it into a cluster mbuf
Xen 3.1.0 public headers [Manuel Bouyer]