Bugzilla Bug 111029

emerge ndiswrapper
Calculating dependencies ...done!
>>> emerge (1 of 1) net-wireless/ndiswrapper-1.2 to /
>>> md5 files   ;-) ndiswrapper-1.2.ebuild
>>> md5 files   ;-) ndiswrapper-1.4.ebuild
>>> md5 files   ;-) files/digest-ndiswrapper-1.2
>>> md5 files   ;-) files/ndiswrapper-1.2-suspend2.patch
>>> md5 files   ;-) files/digest-ndiswrapper-1.4
>>> md5 src_uri ;-) ndiswrapper-1.2.tar.gz
 * Determining the location of the kernel source code
 * Found kernel source directory:
 *     /usr/src/linux
 * Found sources for kernel version:
 *     2.6.14
 * Checking for suitable kernel configuration options:
>>> Unpacking source...
>>> Unpacking ndiswrapper-1.2.tar.gz to /var/tmp/portage/ndiswrapper-1.2/work
 * Applying ndiswrapper-1.2-suspend2.patch ...                                 
                                                                      [ ok ]
 * The only kernels that will work are gentoo-sources, vanilla-sources, and
suspend2-sources.
 * No other kernels are supported. Kernels like the mm kernels will NOT work.
 * Converting ndiswrapper-1.2/driver/Makefile to use M= instead of SUBDIRS= ...
                                                                      [ ok ]
>>> Source unpacked.
cc -Wall -g -DNDISWRAPPER_VERSION=\"1.2\"    -c -o loadndisdriver.o loadndisdriver.c
gcc -o loadndisdriver loadndisdriver.o
 * Preparing ndiswrapper module
make -C //usr/src/linux
M=/var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver \
        NDISWRAPPER_VERSION=1.2 \
        EXTRA_VERSION= modules


*** WARNING: Kernel seems to have 4K size stack option (CONFIG_4KSTACKS)
removed; many Windows drivers will need at least 8K size stacks. You should read
wiki about 4K size stack issue. Don't complain about crashes until you resolve this.


make[1]: Entering directory `/usr/src/linux-2.6.14'
  CC [M]  /var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver/hal.o
  CC [M]  /var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver/iw_ndis.o
  CC [M]  /var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver/loader.o
/var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver/loader.c: In
function `register_devices':
/var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver/loader.c:963:
warning: assignment from incompatible pointer type
  CC [M]  /var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver/misc_funcs.o
  CC [M]  /var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver/ndis.o
  CC [M]  /var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver/ntoskernel.o
/var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver/ndis.c:1721:5:
warning: "LINUX_KERNEL_VERSION" is not defined
  CC [M]  /var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver/pe_linker.o
  CC [M]  /var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver/proc.o
  CC [M]  /var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver/wrapper.o
/var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver/wrapper.c:287:54:
macro "halt" passed 1 arguments, but takes just 0
/var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver/wrapper.c: In
function `miniport_halt':
/var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver/wrapper.c:287:
warning: statement with no effect
  CC [M]  /var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver/usb.o
make[2]: ***
[/var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver/wrapper.o] Error 1
make[2]: *** Waiting for unfinished jobs....
make[1]: ***
[_module_/var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver] Error 2
make[1]: Leaving directory `/usr/src/linux-2.6.14'
make: *** [default] Error 2

!!! ERROR: net-wireless/ndiswrapper-1.2 failed.
!!! Function linux-mod_src_compile, Line 505, Exitcode 2
!!! Unable to make                                  KSRC=//usr/src/linux
KVERS=26 all.
!!! If you need support, post the topmost build error, NOT this status message.


Reproducible: Always
Steps to Reproduce:
emerge vanilla-sources-2.6.14
emerge ndiswrapper-1.2 (ndiswrapper-1.4. is blocked on my system by bug # 110367)


Expected Results:  
emerge should complete successfully

Portage 2.0.51.22-r3 (default-linux/x86/2005.0, gcc-3.3.6, glibc-2.3.5-r2,
2.6.13.4 i686)
=================================================================
System uname: 2.6.13.4 i686 AMD Athlon(tm) XP 3000+
Gentoo Base System version 1.6.13
distcc 2.18.3 i686-pc-linux-gnu (protocols 1 and 2) (default port 3632) [disabled]
dev-lang/python:     2.3.5, 2.4.2
sys-apps/sandbox:    1.2.12
sys-devel/autoconf:  2.13, 2.59-r6
sys-devel/automake:  1.4_p6, 1.5, 1.6.3, 1.7.9-r1, 1.8.5-r3, 1.9.6-r1
sys-devel/binutils:  2.15.92.0.2-r10
sys-devel/libtool:   1.5.20
virtual/os-headers:  2.6.8.1-r1, 2.6.11-r2
ACCEPT_KEYWORDS="x86"
AUTOCLEAN="yes"
CBUILD="i686-pc-linux-gnu"
CFLAGS="-O3 -march=athlon-xp -fomit-frame-pointer -pipe -mmmx -m3dnow -msse
-mfpmath=sse"
CHOST="i686-pc-linux-gnu"
CONFIG_PROTECT="/etc /usr/kde/2/share/config /usr/kde/3.3/env
/usr/kde/3.3/share/config /usr/kde/3.3/shutdown /usr/kde/3.4/env
/usr/kde/3.4/share/config /usr/kde/3.4/shutdown /usr/kde/3/share/config
/usr/lib/X11/xkb /usr/lib/mozilla/defaults/pref /usr/share/config
/var/qmail/control"
CONFIG_PROTECT_MASK="/etc/gconf /etc/terminfo /etc/env.d"
CXXFLAGS="-O3 -march=athlon-xp -fomit-frame-pointer -pipe -mmmx -m3dnow -msse
-mfpmath=sse"
DISTDIR="/usr/portage/distfiles"
FEATURES="autoconfig ccache distlocks fixpackages sfperms strict"
GENTOO_MIRRORS="http://distfiles.gentoo.org
http://distro.ibiblio.org/pub/Linux/distributions/gentoo"
MAKEOPTS="-j3"
PKGDIR="/usr/portage/packages"
PORTAGE_TMPDIR="/var/tmp"
PORTDIR="/usr/portage"
SYNC="rsync://rsync.gentoo.org/gentoo-portage"
USE="x86 3dnow X alsa apm avi berkdb bitmap-fonts cdr crypt cups curl divx
divx4linux eds emboss encode esd fam foomatic foomaticdb fortran gdbm gif gnome
gpm gstreamer gtk gtk2 imagemagick imlib java jpeg junit kerberos ldap libg++
libwww mad matroska mikmod mmx motif mozilla mp3 mp4 mpeg ncurses nls nptl
nptlonly nvidia ogg oggvorbis opengl oss pam pdflib perl png python quicktime
readline samba sdl slang spell sse ssl svga tcpd tiff truetype truetype-fonts
type1-fonts udev unicode vorbis win32codecs xine xinerama xml2 xmms xscreensaver
xv xvid zlib userland_GNU kernel_linux elibc_glibc"
Unset:  ASFLAGS, CTARGET, LANG, LC_ALL, LDFLAGS, LINGUAS, PORTDIR_OVERLAY

I also tried to compile it with gentoo-sources 2.6.14 and it also fails in the
same way.

same here on the gentoo-sources 2.6.14

ok. Generating noise on this bug is going to do nothing. ndiswrapper 1.2 WILL
NOT work on any kernel 2.6.14. Whether it be gentoo-sources or vanilla. You will
need ndiswrapper 1.4. However people complain that 1.4 doesn't work for them,
causes system crashes and general issues. So 1.4 is masked, if you want to
chance it... unmask it and try it. Otherwise stick to an older kernel for now.

Created an attachment (id=72430) [edit]
Patch to run ndiswrapper-1.2 under the 2.6.14 kernel

Actually, here's a patch that has had ndiswrapper 1.2 running fine on
vanilla-sources-2.6.14, and seems much more stable than the new ndiswrapper-1.5
(which I'm just about to post a bug on, having seemingly caused a divide by 0
error in the kernel).

If people would prefer to stick with 1.2, please test out the patch and see if
you have any problems.	It's a very simple renaming of a structure that became
reserved when it was used in the 2.6.14 kernel so there shouldn't be much to
cause regressions or bugs...

I have no time to test this right this moment. I did try 2.6.14 and
ndiswrapper-1.5. While 1.5 does compile and load, I get a timeout on the dhcp
request when net.wlan0 starts, so I had to go back to 2.6.13.5 and ndiswrapper 1.2.

Thanks for the patch! ndiswrapper now compiles cleanly for me with 2.6.14,
however I get the same problem as I did with 1.4 and 1.5
(http://bugs.gentoo.org/show_bug.cgi?id=111754):

$ sudo modprobe ndiswrapper
FATAL: Error inserting ndiswrapper (/lib/modules/2.6.14/misc/ndiswrapper.ko):
Unknown symbol in module, or unknown parameter (see dmesg)

$ dmesg | tail -n 1
ndiswrapper: Unknown symbol kthread_run

I am using vanilla-sources 2.6.14.

mobile: gentoo-sources-2.6.14 is stable, so this should be addressed ASAP. My
suggestions are either make the 1.2 ebuild check the kernel version and die out
on 2.6.14, or apply the change described in comment #4 (it is correct).

patch works for vanilla-sources-2.6.15_rc1

Thanks!! I just tried 2.6.15_rc1 and I'm using ndiswrapper again (finally :)
More info can be provided on request!!

*** Bug 114084 has been marked as a duplicate of this bug. ***

Created an attachment (id=73897) [edit]
modified ebuild that includes the patch

copy the above patch to your $PORTAGE_OVERLAY/net-wireless/files/ folder,
digest and merge this ebuild to have the patch applied automatically

(From update of attachment 73897 [edit])
sorry for the frequent editing, I should watch out for typos...

Created an attachment (id=73900) [edit]
ebuild applying the patch if necessary

supporting kernel 2.6.15* now too... sorry

I just added ndiswrapper-1.7 to three. It is working for me on a stable system.
Adding a patch for 1.2 would demand it go through ~x86 any way so let's see if
1.7 is ready for stable.

Please test with ndiswrapper-1.7.

I've been using ndiswrapper 1.7 on 2.6.15_rc7 and it seems to both compile and
run fine.  I've been running it since December 29th without problems...

As the current stable is not working with 2.6.14 and many people have had
positive experiences with 1.7 I will be marking 1.7 stable next weekend even
though it has not been in ~x86 for very long. If someone from the x86 arch team
has hardware to test take a shot. My hardware works. Feel free to raise
objections too.

(In reply to comment #16)
> As the current stable is not working with 2.6.14 and many people have had
> positive experiences with 1.7 I will be marking 1.7 stable next weekend even
> though it has not been in ~x86 for very long. If someone from the x86 arch team
> has hardware to test take a shot. My hardware works. Feel free to raise
> objections too.

As version 1.7 seems to fix quite a few bugs I think this is the right thing to
do.

If the current stable doesn't work at all, by all means mark one that does
stable as soon as possible.

(In reply to comment #18)
> If the current stable doesn't work at all, by all means mark one that does
> stable as soon as possible.
> 

It works fine with earlier kernels. I would not want to mark something stable
that causes for example kernel panics with 2.6.14 using some other wireless
cards as I only have one model. Well let's see if I will do this already
tomorrow.

I'm using a broadcom 43xx with ndiswrapper 1.7 and I've had no kernel panics
from whatever 2.6.15-rc I started using right up to 2.6.15 which just went
upstream stable.  Hopefully it's of a different chipset to what you've been
testing it with...

(In reply to comment #20)
> I'm using a broadcom 43xx with ndiswrapper 1.7 and I've had no kernel panics
> from whatever 2.6.15-rc I started using right up to 2.6.15 which just went
> upstream stable.  Hopefully it's of a different chipset to what you've been
> testing it with...
> 

It was a theoretical situation. How it performs with 2.6.15 also doesn't matter
from a stabilization point of view that much because stable is 2.6.14. Great to
hear that it is working with 2.6.15 as I haven't had the opportunity to test
more than if it compiles with it yet.

ndiswrapper-1.7 compiles all right with 2.6.14 on my machine. However, when I
use it, network traffic stalls after a couple of minutes, which doesn't happen
to me with ndiswrapper-1.2 on 2.6.13. (1.7 on 2.6.13 stalls as well)

Well it seems ndiswrapper is not the saviour I thought it would be.
bug 117706
So I added 1.2-r1 with the patch from this bug and made 1.2 die if 2.6.14 or
newer is used. Let's see if that can be stabilized.

ndiswrapper-1.7 compiles for me with 2.6.14.x, however, when I use it, network
traffic will stall after a couple of minutes. This happens when I use 1.7 with
both 2.6.13 and 2.6.14. But I'm not sure if that's a ndiswrapper thing or the
windows driver part that't to blame for this. I'm using neta3ab.inf for a
D-link wireless PCI adapter.

The hang up does NOT happen with ndiswrapper-1.2 though, which I'm still
happily using now.

1.2-r1 marked stable on x86. Thanks to everyone who participated in this bug.
Please reopen or make a new bug if 1.2-r1 causes problems for you on 2.6.14.