2.6.14-rc3-nitro1 "Braineating Monsterpenguin"

[ahorn]

got many unresolved symbols with fglrx (6.18.6). the ati-drivers v6.16.14-r5 are working, but since my card is supported first at 6.16.20 (these doesn't work with this patchset, too) i'm a little bit lost. so a patch or a hint or whatever to my unresolved symbols problem would be nice, tiger. or maybe, i use the wrong xorg-version/use-flags (+dlloader,+opengl,xorg-x11-6.8.2-r6)...

i tried this, it helped for loading the fglrx module (the verify_area thing) but the unresolved symbols are still there.

Code: Select all

--- firegl_public.c.orig        2005-08-16 20:13:33.000000000 +0400
+++ firegl_public.c     2005-09-23 17:25:49.000000000 +0400
@@ -121,7 +121,7 @@
 #endif
 
 #ifdef __x86_64__
-#include "asm/ioctl32.h"
+#include "linux/ioctl32.h"
 #if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,2)
 #include "linux/syscalls.h"
 #endif
@@ -1425,7 +1425,7 @@
 
 int ATI_API_CALL __ke_verify_area(int type, const void * addr, unsigned long size)
 {
-    return verify_area(type, addr, size);
+    return access_ok(type, addr, size);
 }
 
 int ATI_API_CALL __ke_get_pci_device_info(__ke_pci_dev_t* dev, __ke_pci_device_info_t *pinfo)

i'm wondering why this patch works for x86. with that patch, there's no verify_area msg in dmesg. and this verify_area is the reason why the fglrx can't be insert. so i'm one step forward.

greetings, ahorn.

[oFooBaro]

ndiswrapper compile fails

Code: Select all

make -C //usr/src/linux M=/var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver \
	NDISWRAPPER_VERSION=1.2 \
	EXTRA_VERSION= modules
make[1]: Entering directory `/usr/src/linux-2.6.14-rc3-nitro1'
  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
/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/ntoskernel.o
  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
make[2]: *** [/var/tmp/portage/ndiswrapper-1.2/work/ndiswrapper-1.2/driver/wrapper.o] Error 1
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-rc3-nitro1'
make: *** [default] Error 2

[Grubshka]

ndiswrapper compile fails

You will need ndiswrapper 1.4, see here to make an ebuild for it.

[m0p]

Well, I'm still on 2.6.12-nitro5, and I want to know, is it really worth upgrading yet? Is 2.6.14 a huge improvement?

[oFooBaro]

ndiswrapper compile fails

You will need ndiswrapper 1.4, see here to make an ebuild for it.

Thanks for your advice.

[thubble]

Whenever I leave the machine for a while (i.e. overnight) and come back to it, I get a weird problem where apps (specifically X apps, and the X server) will lock up for a few seconds. It happens specifically when I'm starting to use an app that hasn't been used for a while, or when switching between windows. There will be about 2 second delay where the whole system locks up, no mouse, keyboard LEDs are stuck, etc. Then everything goes back to normal. I thought it was the swap prefetch patch so I disabled that, same thing.

I'm using the nvidia driver and latest modular xorg-server. I've reversed the patch that was causing problems with nvidia, and there are no relevant lines in dmesg. Pentium 4 with hyperthreading enabled.

[Karl-Franz]

Ok, the old ati-drivers (8.14.13-r5) would not work with this kernel, with the new ones (8.18.whatever) my dualscreen doesn't work.
I suppose following patch would solve it.

Code: Select all

--- firegl_public.c.orig        2005-08-16 20:13:33.000000000 +0400
+++ firegl_public.c     2005-09-23 17:25:49.000000000 +0400
@@ -121,7 +121,7 @@
 #endif
 
 #ifdef __x86_64__
-#include "asm/ioctl32.h"
+#include "linux/ioctl32.h"
 #if LINUX_VERSION_CODE >= KERNEL_VERSION(2,6,2)
 #include "linux/syscalls.h"
 #endif
@@ -1425,7 +1425,7 @@
 
 int ATI_API_CALL __ke_verify_area(int type, const void * addr, unsigned long size)
 {
-    return verify_area(type, addr, size);
+    return access_ok(type, addr, size);
 }
 
 int ATI_API_CALL __ke_get_pci_device_info(__ke_pci_dev_t* dev, __ke_pci_device_info_t *pinfo)

My problem is I don't know how to apply it. I touch a file, enter the text and run try to patch it

Code: Select all

localhost linux # patch -p1 < ../patches/firegl
missing header for unified diff at line 3 of patch
can't find file to patch at input line 3
Perhaps you used the wrong -p or --strip option?
The text leading up to this was:
--------------------------
|--- firegl_public.c.orig   2005-08-16 20:13:33.000000000 +0400
|+++ firegl_public.c   2005-09-23 17:25:49.000000000 +0400
--------------------------
File to patch:

I do not know enough of this to figure out how to do it properly and I'd sure be thankfull if someone would care to explain it to me.

Thanks, KF.

[ahorn]

you have to patch the ati-drivers, not the kernel. look into the portage directory, theres a similar patch which you can modify (ioctl32.patch), add to the ebuild the following: epatch ${FILESDIR}/ioctl32.patch
and then emerge --digest ati-drivers.
BUT why are the 8.14.13-r5 not working at your system. this is the only driver which works here.

anyway, you should read the posts above before you post...

[Karl-Franz]

Thanks alot. It was bitching about the first part so I removed it and now it works like a charm.

Its funny cause the 8.18.6 didn't work with the 2.6.13-r4 kernel (got a res of 1024x768 and only one screen working). And with the 2.6.14-rc4-nitro1 - 8.14.13-r5 combo, startx freezes and I have to reset my computer the windows way.

I don't know what kind of info you like, but I'd sure like to help so if it's anything you need, just walk me through it.

According to the xorg log:

Code: Select all

II) fglrx(0): Kernel Module Version Information:
(II) fglrx(0):     Name: fglrx
(II) fglrx(0):     Version: 8.14.13
(II) fglrx(0):     Date: Jun  8 2005
(II) fglrx(0):     Desc: ATI FireGL DRM kernel module
(WW) fglrx(0): Kernel Module version does *not* match driver.
(EE) fglrx(0): incompatible kernel module detected - HW accelerated OpenGL will not work
(II) fglrx(0): [drm] removed 1 reserved context for kernel
(II) fglrx(0): [drm] unmapping 8192 bytes of SAREA 0xf9c34000 at 0xb7c0f000
(WW) fglrx(0): ***********************************************
(WW) fglrx(0): * DRI initialization failed!                  *
(WW) fglrx(0): * (maybe driver kernel module missing or bad) *
(WW) fglrx(0): * 2D acceleraton available (MMIO)             *
(WW) fglrx(0): * no 3D acceleration available                *
(WW) fglrx(0): ********************************************* *

And it finally breaks down since it's unable to query this kernel module.

/KF

[HecHacker1]

Well, I'm still on 2.6.12-nitro5, and I want to know, is it really worth upgrading yet? Is 2.6.14 a huge improvement?

Me too. I want to upgrade. but ipw2200 doesn't work with anything except the 2.6.12-nitro5 or the genkernel/vanilla kernel. If you enable ipw2200 in this latest nitro for some reason it can't load the firmware.. module or built in. And because this kernel includes 802.11 support built in, you have to use it because the 802.11 package in portage won't compile because of all the 802.11 stuff already in the kernel.

Has anybody got a working ipw2200 with the newest nitro? Or even -mm in general.

[Tiger683]

Code: Select all

(WW) fglrx(0): Kernel Module version does *not* match driver.
(EE) fglrx(0): incompatible kernel module detected - HW accelerated OpenGL will not work

Try doing an opengl-update ati, your kernel module version doesnt match the Xorg opengl driver, thus
no HW acceleration...

I would discourage switching from a stable nitro release to this one (ie. from 2,6,12-nitro5 or 2.6.13-nitro1)

This is a very testing-like release, as many of you have noticed.
until the stable 2.6.14 kernel appears. only the people who want and CAN debug some
problems in it should also use it. Cause: i do not have the hardware to test all the failing drivers on...
Lets hope we will get through this painlessly...

rgds.

T

[HecHacker1]

i found the solution to the problem.. for some reason the ipw2200 drivers in the 2.6.14 release use the 2.2 version of the firmware... Which is two versions old, and one version older than the one in portage.

I enabled the 80211 stack in the kernel (built in) and made the ipw2200 driver as a module. Then you have to manually download the ipw2200-firmware 2.2 and extract it to /lib/firmware

[blk_jack]

Whenever I leave the machine for a while (i.e. overnight) and come back to it, I get a weird problem where apps (specifically X apps, and the X server) will lock up for a few seconds. It happens specifically when I'm starting to use an app that hasn't been used for a while, or when switching between windows. There will be about 2 second delay where the whole system locks up, no mouse, keyboard LEDs are stuck, etc. Then everything goes back to normal. I thought it was the swap prefetch patch so I disabled that, same thing.

I'm using the nvidia driver and latest modular xorg-server. I've reversed the patch that was causing problems with nvidia, and there are no relevant lines in dmesg. Pentium 4 with hyperthreading enabled.

I get the same thing. Memory usage seems fairly high for this release too, a lot higher than older versions of Nitro.

[SchAmane]

I get a weird problem where apps will lock up for a few seconds. It happens specifically when I'm starting to use an app that hasn't been used for a while, or when switching between windows. There will be about 2 second delay where the whole system locks up, no mouse, keyboard LEDs are stuck, etc. Then everything goes back to normal. I thought it was the swap prefetch patch so I disabled that, same thing.

I can agree with that. Enother issue i have with this and 2.6.13.2-nitro1 kernel is slower IO. Mostly on my HDD, but generaly on CD-WRITER and second DVDROM. System mostly hold if i am copying or writing something. Mouse is vely sloooow
One, until i played openttd, whole system get lockup, so i haved to reset my PC.
This kernel is not claimed as stable, so i have no bad hopes about stability, but i would like to know if somebody has IO speed decrase on all >2.6.13 kernels with VIA Chipset like me ?

[HecHacker1]

compared to nitro 2.6.12 rc5 this new version is actually using much LESS memory for me.. However, it just seems that the caching is not being used properly. So yes, less memory usage, but I have to frequently access the hard drive which causes stalls (reiser4, just upgrade to 1.0.5).

I need to take a look at the swap settings cause since I'm running a notebook using the harddrive is not a good thing (and I have a 1GB stick coming soon as an upgrade)

[unz]

i got problems with via kt880, ide0/ide1 resets and dma is disabled ... but seem only warnings, transfers net->hd hd->dvd work at full speed ...

[mbar]

that's it
works here, inserts fine )))

also works with -acid1

[SchAmane]

Yep, thay defenetively have changed something with IO (chipsets support) in last 2 kernels, anybody know ?
DMA option is set for all my CR/DVD Drives but thay feels like thay dont use DMA.
Haved anohter system lock today. After updating world and leaving PC for some hours allone. And it was little bit strange I can still use gnome-terminal (type something), but no new commands, and all other applications seems to no response. I wait few seconds and reset system.

Tiger, are there any other plans to release new 2.6.13.x nitros ? It would be nice "stable nitro" tree

[playfool]

Would you consider adding the adaptive readahead patch which was recently posted on lkml (and covered in this weeks LWN as well btw.) ?

*edit*

Never mind, Andrew Morton included this in the lastest -mm kernel, it will trickle down to vanilla soon enough.

[metalifloyd]

Does this version of Nitro have the staircase scheduler? Unless I'm blind I can't seem to find it.

Also, I tried it out and everything seems to work fine but responsiveness in X is very slow compared to previous Nitro patchsets. That is why I am wondering which CPU scheduler is in use.

[seppe]

Does this version of Nitro have the staircase scheduler? Unless I'm blind I can't seem to find it.

Also, I tried it out and everything seems to work fine but responsiveness in X is very slow compared to previous Nitro patchsets. That is why I am wondering which CPU scheduler is in use.

Yes, the staircase scheduler is in the ck patch. More info at http://kernel.kolivas.org

[metalifloyd]

I always thought Nitro was based on ck.

Hmm, For some reason this nitro is significantly slower than the current ck sources (2.6.13-ck8). Well at least it feels slower. My .config's are pretty similar.

[chichibabin]

Hi,
Compiling ipw2100 and IEEE80211 as modules floods dmesg with the following error:

Code: Select all

eth2 (WE) : Driver using old /proc/net/wireless support, please fix driver !
eth2 (WE) : Driver using old /proc/net/wireless support, please fix driver !
eth2 (WE) : Driver using old /proc/net/wireless support, please fix driver !
eth2 (WE) : Driver using old /proc/net/wireless support, please fix driver !
eth2 (WE) : Driver using old /proc/net/wireless support, please fix driver !
eth2 (WE) : Driver using old /proc/net/wireless support, please fix driver !
eth2 (WE) : Driver using old /proc/net/wireless support, please fix driver !
eth2 (WE) : Driver using old /proc/net/wireless support, please fix driver !
eth2 (WE) : Driver using old /proc/net/wireless support, please fix driver !

My kernel config settings for these 2 modules are as follows:

Code: Select all

# CONFIG_BT_HCIVHCI is not set
CONFIG_IEEE80211=m
# CONFIG_IEEE80211_DEBUG is not set
CONFIG_IEEE80211_CRYPT_WEP=m
CONFIG_IEEE80211_CRYPT_CCMP=m
CONFIG_IEEE80211_CRYPT_TKIP=m

CONFIG_IPW2100=m
# CONFIG_IPW2100_MONITOR is not set
# CONFIG_IPW_DEBUG is not set
# CONFIG_IPW2200 is not set
# CONFIG_AIRO is not set
# CONFIG_HERMES is not set
# CONFIG_ATMEL is not set

Help much appreciated.
Cheers,
Sat

[mbar]

This is connected to recent update of wireless-tools, I've got the same messages but with rt2500.

[sirdilznik]

So far so good AND it seems to use less memory than my previous kernel (2.6.13-rc3-mm1). Great work!