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

Newegg Daily Deals

Reply
 
Thread Tools
Old 04-19-11, 07:30 AM   #13
Deanjo
Registered User
 
Join Date: Aug 2004
Posts: 301
Default Re: VDPAU Feature Set D

Quote:
Originally Posted by zbiggy View Post
According to Nvidia release notes VDPAU C and D are identical. In real world introducing new feature set which is identical to previous one makes no sense. But in marketing world the situation is different - buy VDPAU D because it is just newer than VDPAU C. When you look at GT520 hardware you see GT430 features cut in half and price lower by $10. Who will buy this? - Only people who believe that newer is just better because is more new (OEMs, big shops selling "shop brand" cheap computers). When I read reviews of GT520 I see no good opinion both in international sites and here in Poland. Everyone asks: for who this card was created? Performance of IGP but noise, power consumption, 2 slot cooling size of discrete GPU and price the same as GT430 (its price dropped because GT4xx is now previous generation and shops do shelves cleaning).
Umm sorry but the GT-430 draws a hell of alot more power then the GT-520 and every generation of VDPAU has brought new supported features.

Quote:
Since they started to sell nForce7xx/Geforce 8200 with usb 1.1 hanging on Linux without making any patch to fix this I turned from being Nvidia fan to demanding client and always suspiciously look at dirty Nvidia hands when they try to sell me something.
You know I have a bunch of 8200 based boards and all of them have the USB 1.1 working fine.
Deanjo is offline   Reply With Quote
Old 04-19-11, 11:32 AM   #14
AaronP
NVIDIA Corporation
 
AaronP's Avatar
 
Join Date: Mar 2005
Posts: 2,487
Default Re: VDPAU Feature Set D

Sorry about the confusion. VDPAU feature set D describes improved capabilities in the hardware. However, supporting these features requires software changes that are not yet available. This means that for now, features sets C and D are indeed the same when it comes to the capabilities available to applications.
AaronP is offline   Reply With Quote
Old 04-19-11, 11:53 AM   #15
gradinaruvasile
Registered User
 
gradinaruvasile's Avatar
 
Join Date: May 2008
Posts: 199
Default Re: VDPAU Feature Set D

Quote:
Originally Posted by AaronP View Post
Sorry about the confusion. VDPAU feature set D describes improved capabilities in the hardware. However, supporting these features requires software changes that are not yet available. This means that for now, features sets C and D are indeed the same when it comes to the capabilities available to applications.
And what will be the "improved capabilities in the hardware" when the software changes will be available?

BTW i too have a 8200 board and had not noticed USB 1.1-related hangs.
gradinaruvasile is offline   Reply With Quote
Old 04-19-11, 01:02 PM   #16
zbiggy
Registered User
 
Join Date: Sep 2002
Posts: 623
Default Re: VDPAU Feature Set D

Quote:
Originally Posted by Deanjo View Post
Umm sorry but the GT-430 draws a hell of alot more power then the GT-520 and every generation of VDPAU has brought new supported features.
You will better make use of any recent IGP. Less power less noise and performance close to GT520. For movies no difference.

Quote:
Originally Posted by Deanjo View Post
You know I have a bunch of 8200 based boards and all of them have the USB 1.1 working fine.
Quote:
Originally Posted by gradinaruvasile
BTW i too have a 8200 board and had not noticed USB 1.1-related hangs.
To @gradinaruvasile and @Deanjo
The Nvidia usb 1.1 hanging bug is only present for full speed devices. Keyboards, mice and other low speed hardware are fine.
Full speed usb 1.1 devices like irda dongles, audio cards, ADSL modems will hang after few minutes of use. The only cure for this hang is to hard reset Linux. This usb trap is for Linux exclusively. Windows XP and Solaris are not affected. Nvidia never released kernel patch to fix the issue. You can workaround the trap by using noapic kernel boot parameter.

Do simple test:
Disable usb 2.0 in bios setup. Connect usb harddrive or pendrive. Copy files. It will hang after few minutes.

Now repeat the test on other machine with chipset other than nForce7xx/Geforce82xx. The copy will be fine.

Or connect any full speed usb 1.1 only device which makes intensive transfers over usb without disabling usb 2.0.

I do not blame Nvidia for selling chipset with broken usb ohci controller. I blame them for not providing patch to kernel which will fix it or workaround it.
zbiggy is offline   Reply With Quote
Old 04-19-11, 01:15 PM   #17
zbiggy
Registered User
 
Join Date: Sep 2002
Posts: 623
Default Re: VDPAU Feature Set D

Quote:
Originally Posted by gradinaruvasile View Post
And what will be the "improved capabilities in the hardware" when the software changes will be available?
What about porting these improved capabilities to other GPUs? According to the press GT520 gpu has no architectural changes except being cut in half GT430.
zbiggy is offline   Reply With Quote
Old 04-19-11, 01:32 PM   #18
gradinaruvasile
Registered User
 
gradinaruvasile's Avatar
 
Join Date: May 2008
Posts: 199
Default Re: VDPAU Feature Set D

Quote:
To @gradinaruvasile and @Deanjo
The Nvidia usb 1.1 hanging bug is only present for full speed devices. Keyboards, mice and other low speed hardware are fine.
Full speed usb 1.1 devices like irda dongles, audio cards, ADSL modems will hang after few minutes of use. The only cure for this hang is to hard reset Linux. This usb trap is for Linux exclusively. Windows XP and Solaris are not affected. Nvidia never released kernel patch to fix the issue. You can workaround the trap by using noapic kernel boot parameter.

Do simple test:
Disable usb 2.0 in bios setup. Connect usb harddrive or pendrive. Copy files. It will hang after few minutes.

Now repeat the test on other machine with chipset other than nForce7xx/Geforce82xx. The copy will be fine.

Or connect any full speed usb 1.1 only device which makes intensive transfers over usb without disabling usb 2.0.

I do not blame Nvidia for selling chipset with broken usb ohci controller. I blame them for not providing patch to kernel which will fix it or workaround it.
So far i tried: USB pendrives, 3G modems (3-4 models), cell phones (USB transfer mode), cameras, SD/MMC card readers, webcams. No hangs.
Anyway, i dont deny the fact that there is such a bug, but i did not encounter it.

I have an ASUS M3N78-VM mobo with MCP78/8200 chipset. Using Ubuntu 9.10/Debian Squeeze/Testing with the (Ubuntu 9.10 default) i think 2.6.28, Debian 2.6.32 and 2.6.38 kernels.
As far as i know "full speed" for USB means 2.0. And if it works well in Windows/Solaris, why people say that the controller is broken?
gradinaruvasile is offline   Reply With Quote
Old 04-19-11, 05:41 PM   #19
Deanjo
Registered User
 
Join Date: Aug 2004
Posts: 301
Default Re: VDPAU Feature Set D

Quote:
Originally Posted by zbiggy View Post
You will better make use of any recent IGP. Less power less noise and performance close to GT520. For movies no difference.
You would think that but alas that is not true. Intel IGP's for example still cannot deliver a true ntsc film rate( which means frame duplication) , only accelerate h264 and mpeg2. AMD's IGP's are absolutely horrid in linux (vaapi dxVA wrapper still has ALOT of growing pains) when it comes to hardware acceleration as well. Also a GT520 is far cheaper then replacing an entire system.


Quote:

To @gradinaruvasile and @Deanjo
The Nvidia usb 1.1 hanging bug is only present for full speed devices. Keyboards, mice and other low speed hardware are fine.
Full speed usb 1.1 devices like irda dongles, audio cards, ADSL modems will hang after few minutes of use.
I am well aware what "Full speed usb 1.1 devices" are. No issues here with usb 1.1 audio cards, older usb storage (hello ipod) or even an old USB 1.1 wireless dongle. No work arounds here needed. Works as advertised.

Quote:
Do simple test:
Disable usb 2.0 in bios setup. Connect usb harddrive or pendrive. Copy files. It will hang after few minutes.

Now repeat the test on other machine with chipset other than nForce7xx/Geforce82xx. The copy will be fine.

Or connect any full speed usb 1.1 only device which makes intensive transfers over usb without disabling usb 2.0.

I do not blame Nvidia for selling chipset with broken usb ohci controller. I blame them for not providing patch to kernel which will fix it or workaround it.
As I said can do these no problem.
Deanjo is offline   Reply With Quote
Old 04-19-11, 05:45 PM   #20
Deanjo
Registered User
 
Join Date: Aug 2004
Posts: 301
Default Re: VDPAU Feature Set D

Quote:
Originally Posted by gradinaruvasile View Post
So far i tried: USB pendrives, 3G modems (3-4 models), cell phones (USB transfer mode), cameras, SD/MMC card readers, webcams. No hangs.
Anyway, i dont deny the fact that there is such a bug, but i did not encounter it.

I have an ASUS M3N78-VM mobo with MCP78/8200 chipset.
Same motherboards here as well as a M3N-HT and a couple of M4N78-VM that do not have any USB 1.1 issues.
Deanjo is offline   Reply With Quote

Old 04-19-11, 05:54 PM   #21
Deanjo
Registered User
 
Join Date: Aug 2004
Posts: 301
Default Re: VDPAU Feature Set D

Quote:
Originally Posted by zbiggy View Post
What about porting these improved capabilities to other GPUs? According to the press GT520 gpu has no architectural changes except being cut in half GT430.
The GT-430 is GF-118, the GT-520 is a GF-119. Although derived from the same fermi architecture it does not mean that the video decode engine remained the same. The GF 119 is not simply a case of just disabling 8 ROPs on a bad GF-118 or another example of the GF100 having part of the chip neutered.
Deanjo is offline   Reply With Quote
Old 04-19-11, 08:17 PM   #22
primerib
Registered User
 
Join Date: Dec 2008
Posts: 128
Default Re: VDPAU Feature Set D

Quote:
The Nvidia usb 1.1 hanging bug is only present for full speed devices. Keyboards, mice and other low speed hardware are fine.
Full speed usb 1.1 devices like irda dongles, audio cards, ADSL modems will hang after few minutes of use.
You can include me in the group of people who don't have this problems. Maybe it's specific to a certain usb controller. Or maybe crappy hardware. I don't know. However, when this many people aren't experiencing the problem, it's probably wise to look for the root cause elsewhere. Maybe Nvidia never fixed it because it's not an Nvidia problem.
primerib is offline   Reply With Quote
Old 04-19-11, 09:42 PM   #23
Deanjo
Registered User
 
Join Date: Aug 2004
Posts: 301
Default Re: VDPAU Feature Set D

Quote:
Originally Posted by AaronP View Post
VDPAU feature set D describes improved capabilities in the hardware. However, supporting these features requires software changes that are not yet available.

And these planned additions are..........?
Deanjo is offline   Reply With Quote
Old 04-20-11, 02:33 PM   #24
zbiggy
Registered User
 
Join Date: Sep 2002
Posts: 623
Default Re: VDPAU Feature Set D

Quote:
Originally Posted by gradinaruvasile View Post
So far i tried: USB pendrives, 3G modems (3-4 models), cell phones (USB transfer mode), cameras, SD/MMC card readers, webcams. No hangs.
Anyway, i dont deny the fact that there is such a bug, but i did not encounter it.

I have an ASUS M3N78-VM mobo with MCP78/8200 chipset. Using Ubuntu 9.10/Debian Squeeze/Testing with the (Ubuntu 9.10 default) i think 2.6.28, Debian 2.6.32 and 2.6.38 kernels.
As far as i know "full speed" for USB means 2.0. And if it works well in Windows/Solaris, why people say that the controller is broken?
I have AsRock K10N78FullHD-hSLI R3.0 - 2 pieces (one bought in 2008, second one bought in 2009 in another shop to prove this is not only one faulty). I tried windows xp sp3, solaris 10, ubuntu, suse, fedora live cds 32 and 64bit, RHEL5, CentoOS5. Windows and Solaris do not fail because they use slow, level triggered APIC. Linux as the only one uses new fast fasteoi mode APIC. This may be the reason and that is why using noapic kernel boot parameter workarounds the bug.

Thanks to Nvidia I'm educated usb user now. USB 1.1 has two modes of working depending what device is plugged in: slow and full speed. USB 2.0 is hi speed. That is why usb 1.1 keyboard and mice works and hi speed devices work like hard disks or video grabbers.

Geforce 8200 or nForce 7xxx mainboards with broken Nvidia usb 1.1 full speed mode:
Pegatron M2N78-LA
AsRock K10N78FullHD-hSLI R3.0 (2 pieces bought and failed)
Shuttle SN78SH7 (~ 25 pieces bought and failed)
ASUS M3N78 PRO
Asus M3N78-VM <-- the same as yours so you made test incorrectly (probably you did not disabled usb 2.0 in bios setup before testing and performed the test using usb 2.0 which is fine).

Lets compare working and broken Nvidia chipsets: lspci -nn | grep OHCI
broken:
Code:
00:02.0 USB Controller [0c03]: nVidia Corporation MCP78S [GeForce 8200] OHCI USB 1.1 Controller [10de:077b] (rev a1)
00:04.0 USB Controller [0c03]: nVidia Corporation MCP78S [GeForce 8200] OHCI USB 1.1 Controller [10de:077d] (rev a1)
zbiggy 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 12:46 AM.


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