nV News Forums

 
 

nV News Forums (http://www.nvnews.net/vbulletin/index.php)
-   NVIDIA Linux (http://www.nvnews.net/vbulletin/forumdisplay.php?f=14)
-   -   Kernel 3.6-rc1 breaks 304.30 compile.. (http://www.nvnews.net/vbulletin/showthread.php?t=187620)

rg2003 08-03-12 06:20 AM

Kernel 3.6-rc1 breaks 304.30 compile..
 
Just updated to kernel 3.6-rc1 (from kernel.org git) and the latest driver (304.30) does not compile:

........ from /home/rgadsdon/NVIDIA-Linux-x86_64-304.30/kernel/nv-acpi.c:15:
/usr/src/linux-3.6-rc1/arch/x86/include/asm/uaccess_64.h: In function ‘copy_from_user':
/usr/src/linux-3.6-rc1/arch/x86/include/asm/uaccess_64.h:62:41: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]
/home/rgadsdon/NVIDIA-Linux-x86_64-304.30/kernel/nv-acpi.c: In function ‘nv_acpi_remove':
/home/rgadsdon/NVIDIA-Linux-x86_64-304.30/kernel/nv-acpi.c:303:9: error: too many arguments to function ‘acpi_os_wait_events_complete'
In file included from include/acpi/acpi.h:63:0,
from /home/rgadsdon/NVIDIA-Linux-x86_64-304.30/kernel/nv-linux.h:274,
from /home/rgadsdon/NVIDIA-Linux-x86_64-304.30/kernel/nv-acpi.c:15:
include/acpi/acpiosxf.h:208:6: note: declared here
make[4]: *** [/home/rgadsdon/NVIDIA-Linux-x86_64-304.30/kernel/nv-acpi.o] Error
1
make[3]: *** [_module_/home/rgadsdon/NVIDIA-Linux-x86_64-304.30/kernel] Error 2
make[2]: *** [sub-make] Error 2
NVIDIA: left KBUILD.
nvidia.ko failed to build!
make[1]: *** [module] Error 1
make: *** [install] Error 2

leigh123linux 08-03-12 07:11 AM

Re: Kernel 3.6-rc1 breaks 304.30 compile..
 
http://cvs.rpmfusion.org/viewvc/rpms...ee&view=markup

rg2003 08-03-12 08:46 AM

Re: Kernel 3.6-rc1 breaks 304.30 compile..
 
@leigh123linux, Thanks!.. Applied the patch, and driver compiles - and runs - OK..

Robert Gadsdon
rglinuxtech.com

siddly 08-03-12 08:47 AM

Re: Kernel 3.6-rc1 breaks 304.30 compile..
 
Quote:

Originally Posted by rg2003 (Post 2570724)
Just updated to kernel 3.6-rc1 (from kernel.org git) and the latest driver (304.30) does not compile:

........ from /home/rgadsdon/NVIDIA-Linux-x86_64-304.30/kernel/nv-acpi.c:15:
/usr/src/linux-3.6-rc1/arch/x86/include/asm/uaccess_64.h: In function ‘copy_from_user':
/usr/src/linux-3.6-rc1/arch/x86/include/asm/uaccess_64.h:62:41: warning: comparison between signed and unsigned integer expressions [-Wsign-compare]
/home/rgadsdon/NVIDIA-Linux-x86_64-304.30/kernel/nv-acpi.c: In function ‘nv_acpi_remove':
/home/rgadsdon/NVIDIA-Linux-x86_64-304.30/kernel/nv-acpi.c:303:9: error: too many arguments to function ‘acpi_os_wait_events_complete'
In file included from include/acpi/acpi.h:63:0,
from /home/rgadsdon/NVIDIA-Linux-x86_64-304.30/kernel/nv-linux.h:274,
from /home/rgadsdon/NVIDIA-Linux-x86_64-304.30/kernel/nv-acpi.c:15:
include/acpi/acpiosxf.h:208:6: note: declared here
make[4]: *** [/home/rgadsdon/NVIDIA-Linux-x86_64-304.30/kernel/nv-acpi.o] Error
1
make[3]: *** [_module_/home/rgadsdon/NVIDIA-Linux-x86_64-304.30/kernel] Error 2
make[2]: *** [sub-make] Error 2
NVIDIA: left KBUILD.
nvidia.ko failed to build!
make[1]: *** [module] Error 1
make: *** [install] Error 2

Edit NVIDIA-Linux-x86_64-304.30/kernel/nv-acpi.c by removing the NULL in brackets.

Fix
---
- acpi_os_wait_events_complete(NULL);
+ acpi_os_wait_events_complete();

rockob 08-13-12 09:13 PM

Re: Kernel 3.6-rc1 breaks 304.30 compile..
 
This bug is still present in 304.37.

leigh123linux 08-13-12 09:43 PM

Re: Kernel 3.6-rc1 breaks 304.30 compile..
 
Quote:

Originally Posted by rockob (Post 2571041)
This bug is still present in 304.37.

This isn't a bug IMO, why should nvidia support an unreleased kernel?

siddly 08-13-12 10:53 PM

Re: Kernel 3.6-rc1 breaks 304.30 compile..
 
Quote:

Originally Posted by leigh123linux (Post 2571046)
This isn't a bug IMO, why should nvidia support an unreleased kernel?

It's not strictly a bug, I would call it a heads up as kernel API's are subject to frequent change they do break the driver and that has to happen with software maintained outside the kernel tree that depends on kernel API's.
Reports such as these are in the hope that the developers' work is made lighter and in appreciation of the work they do to support NVidia products, products that are critical and valued.
1. Some people test kernels for such problems and report back.
2. We have no way of knowing how closely the developers in their daily grind are following API changes.
3. It's better to be proactive than wait for a "stable" kernel release only then to find there is a problem.
4. There are also circumstances where people need an unreleased kernel to fix a problem or support for a device that's not supported in a released kernel.
As Linus says, the one constant in Linux is change.
In many instances we are fortunate in Linux, e.g when we had an urgent need for UAC2 support it was written in to 2.6.35 and has since been maintained within the kernel tree ever since. Another OS still doesn't support UAC2 and it's proponents can only shout at Linux developers for not writing a driver for their OS, reason being their OS provider will not even do them the courtesy of a response should they dare to raise the problem where it rightly should be raised.

We like to think NVidia is more responsive to it's customers' needs and I've found it to be so over many years.


All times are GMT -5. The time now is 07:04 PM.

Powered by vBulletin® Version 3.7.1
Copyright ©2000 - 2014, Jelsoft Enterprises Ltd.
Copyright ©1998 - 2014, nV News.