Gentoo Archives: gentoo-user

From: "S. Schwartz" <sigi.schwartz@×××××××××××××.de>
To: gentoo-user@l.g.o
Subject: [gentoo-user] nvidia-kernel vs. gentoo-sources
Date: Mon, 01 May 2006 23:22:46
Message-Id: 44569349.8000900@habmalnefrage.de
1 Hi,
2
3 I'm having a problem using nvidia-kernel-1.0.6629-r5 after the my recent
4 kernel-update from gentoo-sources-2.6.15-r1 to gentoo-sources-2.6.16-r3.
5 The few things, that are compiled when emerging nvidia-kernel, compile
6 successfully but emerge complains about missing symboles
7 (remap_page_range and pci_find_class).
8
9 Via Google I found out that nvidia-kernel-1.0.6111-rX had similar
10 problems with pci_find_class, which was fixed eventually (that function
11 seems to have been renamed to pci_get_class). That one seems to be back
12 along with a new friend (the function remap_page_range has been renamed
13 to remap_pfn_range and expects a different number of parameters I
14 think). During emerge the script conftest.sh from the
15 nvidia-kernel-package should take care of at least remap_page_range, but
16 calling this script fails. I found heavy differences between
17 conftest.sh-scripts from the currend nvidia-kernel and from older
18 versions. To me it seems, the patching in the current version is not
19 done properly. Still I'm too unexperienced with patching in order to do
20 it myself.
21
22 Does anyone experience same problems or know about a solution to that? I
23 tried all newer versions of nvidia-kernel. Some didn't compile or didn't
24 load afterwards while one of them did both successfully but seriously
25 crashed when starting X (luckily remote-login still worked).
26
27 Sigi
28
29 --
30 gentoo-user@g.o mailing list

Replies