Using vanilla-sources 2.4.21, then make dep, then emerge nvidia-kernel, then going back to make menuconfig and selecting []NVIDIA chipset support (under /dev/agpgart) results in failure of make bzImage (after another make dep of course) at: gcc -D__KERNEL__ -I/usr/src/linux-2.4.21/include -Wall -Wstrict-prototypes -Wno-trigraphs -O2 -fno-strict-aliasing -fno-common -fomit-frame-pointer -pipe -mpreferred-stack-boundary=2 -march=athlon -nostdinc -iwithprefix include -DKBUILD_BASENAME=agpgart_be -DEXPORT_SYMTAB -c agpgart_be.c agpgart_be.c: In function `agp_generic_create_gatt_table': agpgart_be.c:583: warning: assignment from incompatible pointer type agpgart_be.c: In function `nvidia_generic_setup': agpgart_be.c:4304: structure has no member named `num_of_masks' agpgart_be.c: At top level: agpgart_be.c:674: warning: `agp_generic_insert_memory' defined but not used agpgart_be.c:736: warning: `agp_generic_remove_memory' defined but not used make[4]: *** [agpgart_be.o] Error 1 make[4]: Leaving directory `/usr/src/linux-2.4.21/drivers/char/agp' make[3]: *** [first_rule] Error 2 make[3]: Leaving directory `/usr/src/linux-2.4.21/drivers/char/agp' make[2]: *** [_subdir_agp] Error 2 make[2]: Leaving directory `/usr/src/linux-2.4.21/drivers/char' make[1]: *** [_subdir_char] Error 2 make[1]: Leaving directory `/usr/src/linux-2.4.21/drivers' make: *** [_dir_drivers] Error 2 Note getting the patch from Nvidia, installing it manually, then trying to compile with that option selected also fails - so this may be incompatibility between the patch and 2.4.21 - even though it installs cleanly. (It looks like the 2.6 kernels support Nvidia AGP - but I also need the nforce modules for NIC and sound, and these don't appear to be available for 2.6 yet.) Is NVIDIA chipset support not supposed to be selected anyway? Does nvidia-kernel just assemble a module for AGP outside of the kernel build process? If so, this should be documented. The main Gentoo installation doc suggests that "extra configuration" is needed for nvidia-kernel, but doesn't specify what, so checking off the obvious new option in the kernel menu seemed suggested. Reproducible: Always Steps to Reproduce: 1. 2. 3. Portage 2.0.48-r5 (default-x86-1.4, gcc-3.2.3, glibc-2.3.2-r1) ================================================================= System uname: 2.4.21 i686 AMD Athlon(tm) XP 2500+ GENTOO_MIRRORS="http://gentoo.oregonstate.edu http://distro.ibiblio.org/pub/Linux/distributions/gentoo" CONFIG_PROTECT="/usr/kde/2/share/config /etc /var/qmail/control /var/bind /usr/X11R6/lib/X11/xkb /usr/kde/3.1/share/config /usr/kde/3/share/config /usr/share/config" CONFIG_PROTECT_MASK="/etc/gconf /etc/env.d" PORTDIR="/usr/portage" DISTDIR="/usr/portage/distfiles" PKGDIR="/usr/portage/packages" PORTAGE_TMPDIR="/var/tmp" PORTDIR_OVERLAY="" USE="x86 oss 3dnow apm avi crypt cups encode foomaticdb gif jpeg kde libg++ mad mikmod mmx mpeg ncurses nls pdflib png quicktime spell truetype xml2 xmms xv zlib gdbm berkdb slang readline arts tetex svga tcltk guile mysql X sdl gpm tcpd pam libwww ssl perl python esd imlib oggvorbis gtk qt motif opengl mozilla cdr -java -gnome mbox -apache2" COMPILER="gcc3" CHOST="i686-pc-linux-gnu" CFLAGS="-mcpu=athlon-xp -O3 -pipe" CXXFLAGS="-O2 -mcpu=i686 -pipe" ACCEPT_KEYWORDS="x86" MAKEOPTS="-j2" AUTOCLEAN="yes" SYNC="rsync://rsync.gentoo.org/gentoo-portage" FEATURES="sandbox ccache"
emerging nvidia-glx got past this problem to the next one (an error about a variable named "name"), which nvidia-kernel-1.0.4496-r1.ebuild has solved.
so can this bug be closed then?
err - sorry have too many puzzles before me at once. The "solution" I just posted was to a subsequent puzzle where it failed to build even after removing agpgart from selection in the kernel menus. So I think the "right" way to do this is to _not_ have agpgart or NVIDIA selected in the kernel menus, and then to use the more recent masked version if using a 2.6 kernel. But I haven't fully tested that. Both of these were from hints I found in the forums. So, eh, the original report may still be pointing to a problem, although a work-around may be at hand.
The nvidia-kernel side should be fixed. Please try to keep the bugs seperate, iow, file a bug for the kernel failing, and another for nvidia-kernel failing. Anyhow, If I understand correctly, he had an issue with the recent drop in dev.name variable for 2.6.0-test3-bk5, but also a failing in 2.4.21 not compiling the nvidia agp module. The nvidia-kernel issue is fixed, but I have no idea about the kernel side. Is that a gentoo kernel, or vanilla sources ?
Vanilla sources - so that issue's a kernel problem.
Martin, You may also want to look at bug # 27412. I'm off the 2.4.x kernels now and I don't know how much of the new ACPI stuff may have been back ported. Best regards, Guy
Nobody kernel side seems to take this. If you have a problem with non vanilla sources (meaning a kernel you 'emerge'd, please file a bug against that kernel. That is what they are there for - to fix issues with vanilla, plus enhancements.