Go Back   nV News Forums > Linux Support Forums > NVIDIA Linux

Newegg Daily Deals

Reply
 
Thread Tools
Old 02-14-07, 04:32 PM   #37
shutdown
Registered User
 
Join Date: Nov 2005
Posts: 45
Default Re: Unified Xen and real-time preemption patch for 1.0-9625

Does the kernel print any messages when interrupting the boot process? I can't think of any config options of the kernel that cause such behaviour, but there are too many to be sure that there isn't one of that kind.

Applying -rt should work fine if you're doing the following:
Code:
$ cd /usr/src/linux
$ patch -p1 < /path/to/rt-patch
Replace "/usr/src/linux" with your kernel source dir if you did not link it to /usr/src/linux already and replace "/patch/to/rt-patch" with the path to the rt-patch.
I guess you just missed the "<" after the -p1 option for patch. This makes the shell read the given file and pass the output to the given command.

Shutdown
shutdown is offline   Reply With Quote
Old 02-14-07, 05:05 PM   #38
floogy
Registered User
 
Join Date: Aug 2006
Posts: 34
Default Re: Unified Xen and real-time preemption patch for 1.0-9625

The kernel doesn't print anything, it just stops booting, and wait for a return, to keep up booting again. This is maybe because of boot splashes that are not accessible ?? Hard to say...

Regarding patching, I meant the patch for the nvidia driver, to compile against the rt-patched kernel, because there appeared several times the version number 9625, and other bizarre names and pathes (but I don't know much about patching and diffs):
http://www.nvnews.net/vbulletin/atta...3&d=1171124108
floogy is offline   Reply With Quote
Old 02-14-07, 05:24 PM   #39
shutdown
Registered User
 
Join Date: Nov 2005
Posts: 45
Default Re: Unified Xen and real-time preemption patch for 1.0-9625

Quote:
Originally Posted by floogy
Regarding patching, I meant the patch for the nvidia driver, to compile against the rt-patched kernel, because there appeared several times the version number 9625, and other bizarre names and pathes (but I don't know much about patching and diffs):
http://www.nvnews.net/vbulletin/atta...3&d=1171124108
You mean that in the patch file the number 9625 appears multiple times? Don't worry about that, these are part of paths to files to patch. Actually the diffs contain 2 paths for each file, an "old" file and a "new" file and the diff itself holds the information of the difference between those.
So if you create a diff file you copy the source of the code you want to patch to another location, modify it there and then run a diff to compare the old and the new version.
Because the patch for 1.0-9625 worked for 1.0-9746 and just needed changes of the line numbers and one little addition I simply used the old patch and modified it, but left the filenames untouched because they don't matter in the end as you will see:

To patch your nVidia driver you have to do the following:
Code:
$ ./NVIDIA-Linux-x86_64-1.0-9746-pkg2.run -x
$ cd cd NVIDIA-Linux-x86_64-1.0-9746-pkg2/usr/src/nv
$ patch -p1 < /path/to/patch
You probably did not change your working directory to the usr/src/nv directory of the extracted nVidia driver which will render the patch utility unable to locate the files to patch in this situation.
I could have created a completely new patch which can be run from the extracted nVidia driver directory itself, but because the patch for 1.0-9625 needed to be run from usr/src/nv I decided to only modify the existing patch and don't change the way one has to apply the patch.
Actually I haven't seen any patch for the nvidia driver which was NOT run from inside this directory lately, so I thought "why change something that is common?"...

Shutdown
shutdown is offline   Reply With Quote
Old 02-14-07, 05:36 PM   #40
floogy
Registered User
 
Join Date: Aug 2006
Posts: 34
Default Re: Unified Xen and real-time preemption patch for 1.0-9625

Thank you for the good explanation. Now I get a picture of what's going on while patching sources ;-) especially nvidia sources.
floogy is offline   Reply With Quote
Old 02-17-07, 07:49 AM   #41
shutdown
Registered User
 
Join Date: Nov 2005
Posts: 45
Default Re: Unified Xen and real-time preemption patch for 1.0-9625

2.6.20-rt8 seems to be pretty functional and stable again, at least for me - no problems booting it and it's running without complaining about anything for about half an hour now on my box. dmesg looks good, so I'd say it's worth a try for everyone who had problems.

Shutdown
shutdown is offline   Reply With Quote
Old 02-19-07, 02:59 AM   #42
floogy
Registered User
 
Join Date: Aug 2006
Posts: 34
Default Re: Unified Xen and real-time preemption patch for 1.0-9625

Hi shutdown,

This is a little OT: I tried a lot of .config options, but I can' find the option that makes the kernel stop on loading several modules, and waiting for a RETURN to keep up booting.

Any ideas? Or even helpful search words to google for it ?
floogy is offline   Reply With Quote
Old 02-19-07, 04:10 AM   #43
shutdown
Registered User
 
Join Date: Nov 2005
Posts: 45
Default Re: Unified Xen and real-time preemption patch for 1.0-9625

Have you tried the same config for a kernel without the -rt patchset and does it behave in the same way?
If we know that it's not -rt related there must be another config option causing your problems, but if not it's caused by -rt. I think we should figure this out first to know where to look for the problem.

Shutdown
shutdown is offline   Reply With Quote
Old 02-19-07, 05:32 AM   #44
JaXXoN
Registered User
 
Join Date: Jul 2005
Location: Munich
Posts: 910
Default Re: Unified Xen and real-time preemption patch for 1.0-9625

Just to let you know: i'm sucessfully running shutdown's patch for 9746 with
2.6.20-rt5 on an Athlon X2. Maybe the attached config file gives you a clue
what could be the problem ...

regards

Bernhard
Attached Files
File Type: zip config-2.6.20-rt5.zip (19.6 KB, 202 views)
JaXXoN is offline   Reply With Quote

Old 02-19-07, 06:21 AM   #45
floogy
Registered User
 
Join Date: Aug 2006
Posts: 34
Default Re: Unified Xen and real-time preemption patch for 1.0-9625

Thank You JaXXoN, I'll try a diff between the two .config's.
Hopefully I'll find the responsible option.
I also think, that this is unrelated to the patch, that applied fine here.
Sad enough, but I messed up the .config myself .
floogy is offline   Reply With Quote
Old 02-21-07, 04:58 AM   #46
shutdown
Registered User
 
Join Date: Nov 2005
Posts: 45
Default Re: Unified Xen and real-time preemption patch for 1.0-9625

@ floogy: Have you had success finding the cause of your problems yet?

@ everyone: 2.6.21-rc1 is out and the dyntick and high-res-timer patches are really included! So -rt will become a lot less experimental and buggy within the next 1 or 2 months I hope.

Edit: After applying the patch I cannot select hrtimers nor dynticks, can't find it anywhere in the config...if anyone can tell me where the hell the kconfig option has been moved or what it depends on I'd be quite happy...

Shutdown
shutdown is offline   Reply With Quote
Old 02-22-07, 05:35 AM   #47
floogy
Registered User
 
Join Date: Aug 2006
Posts: 34
Default Re: Unified Xen and real-time preemption patch for 1.0-9625

Hi shutdown,

No I didn't had luck. Maybe there are now less breaks (?):

Code:
$ uname -a
Linux ubuntu 2.6.20-rt5-5 #1 SMP PREEMPT Mon Feb 19 20:29:13 CET 2007 x86_64 GNU/Linux
$ dmesg
.000000] On node 0 totalpages: 262031
[    0.000000]   DMA zone: 96 pages used for memmap

[...] 1. break 31 - 60

[   31.020026] NFORCE-CK804: 0000:00:06.0 (rev a2) UDMA133 controller
[   31.020130]     ide0: BM-DMA at 0xf000-0xf007, BIOS settings: hda:DMA, hdb:DMA
[   31.020257]     ide1: BM-DMA at 0xf008-0xf00f, BIOS settings: hdc:DMA, hdd:DMA
[   31.020382] Probing IDE interface ide0...
[   60.325407] hda: HL-DT-ST DVDRAM GSA-4167B, ATAPI CD/DVD-ROM drive
[   60.632414] ide0 at 0x1f0-0x1f7,0x3f6 on irq 14
[   60.632539] Probing IDE interface ide1...
[   61.304420] hdc: -DT-ST DVDRAM GSA-4167B, ATAPI CD/DVD-ROM drive

[...] 2.+3. brake 63 - 88 - 112

[   63.373569] ohci_hcd 0000:00:02.0: irq 20, io mem 0xd2004000
[   63.429484] usb usb2: configuration #1 chosen from 1 choice
[   63.433065] hub 2-0:1.0: USB hub found
[   63.436640] hub 2-0:1.0: 10 ports detected
[   88.889773] usb 1-1: new high speed USB device using ehci_hcd and address 2
[  112.264760] usb 1-1: configuration #1 chosen from 1 choice
[  112.273201] usbcore: registered new interface driver libusual
[  112.282486] Initializing USB Mass Storage driver...
[  112.286108] scsi4 : SCSI emulation for USB Mass Storage devices
[  112.289830] usb-storage: device found at 2

[...] 4. brake 117 - 122

[  115.764580] Attempting manual resume
[  116.554243] kjournald starting.  Commit interval 5 seconds
[  116.558008] EXT3-fs: mounted filesystem with ordered data mode.
[  117.291940] scsi 4:0:0:0: Direct-Access     Lexmark  USB Mass Storage 1.00 PQ: 0 ANSI: 0
[  117.297710] sd 4:0:0:0: Attached scsi removable disk sdb
[  117.301708] usb-storage: device scan complete
[  122.620541] PCI: Enabling device 0000:00:04.0 (0000 -> 0003)
[  122.624318] ACPI: PCI Interrupt Link [APCJ] enabled at IRQ 22
[  122.627656] ACPI: PCI Interrupt 0000:00:04.0[A] -> Link [APCJ] -> GSI 22 (level, low) -> IRQ 22
[  122.631040] PCI: Setting latency timer of device 0000:00:04.0 to 64
[  122.908321] sd 2:0:0:0: Attached scsi generic sg0 type 0

[...] normal booting

[  184.219580] ttyS0: LSR safety check engaged!
[  184.910353] NET: Registered protocol family 17
[  184.930267] device eth0 entered promiscuous mode
[  184.930443] audit(1172095365.066:2): dev=eth0 prom=256 old_prom=0 auid=4294967295
[  185.030784] device eth0 left promiscuous mode
[  185.030951] audit(1172095365.066:3): dev=eth0 prom=0 old_prom=256 auid=4294967295
[  199.412217] UDF-fs: No VRS found
[  200.381664] UDF-fs: No VRS found
[  200.503171] ISO 9660 Extensions: Microsoft Joliet Level 3
[  200.503977] ISOFS: changing to secondary root

End of dmesg
floogy is offline   Reply With Quote
Old 05-05-07, 09:35 PM   #48
Nexus Scorpion
Registered User
 
Join Date: Apr 2005
Posts: 4
Default Re: Unified Xen and real-time preemption patch for 1.0-9625

For anyone who's interested, I have (trivially) backported this patch to version 1.0-8776. The patch is attached. I have tested it with Xen, though not with RTLinux.
Attached Files
File Type: txt patch-nv-1.0-8776-xenrt.txt (4.0 KB, 785 views)
Nexus Scorpion is offline   Reply With Quote
Reply


Thread Tools

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Forum Jump


All times are GMT -5. The time now is 12:48 PM.


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