Topic: [SLVD] G\A\CX_95952.cat is a one-byte-file

The mentioned file is not a valid certificate, because of this the driver is rejected, I guess, and the built-in driver is used. I have disabled Driver Signing in WinXP/SP3 via dialogue box.

[2011/11/03 11:05:18 552.1298 Driver Install]
#-019 Searching for hardware ID(s): pci\ven_1002&dev_4151&subsys_7c37174b&rev_00,pci\ven_1002&dev_4151&subsys_7c37174b,pci\ven_1002&dev_4151&cc_030000,pci\ven_1002&dev_4151&cc_0300
#-018 Searching for compatible ID(s): pci\ven_1002&dev_4151&rev_00,pci\ven_1002&dev_4151,pci\ven_1002&cc_030000,pci\ven_1002&cc_0300,pci\ven_1002,pci\cc_030000,pci\cc_0300
#-199 Executing "C:\WINXP\system32\setup.exe" with command line: setup.exe -newsetup -mini
#I022 Found "PCI\VEN_1002&DEV_4151" in c:\d\g\a\cx_95952.inf; Device: "RADEON 9600 SERIES"; Driver: "RADEON 9600 SERIES"; Provider: "ATI Technologies Inc."; Mfg: "ATI Technologies Inc."; Section name: "ati2mtag_RV350".
#I087 Driver node not trusted, rank changed from 0x00002001 to 0x0000a001.
#I023 Actual install section: [ati2mtag_RV350]. Rank: 0x0000a001. Effective driver date: 02/10/2010.
#I022 Found "PCI\VEN_1002&DEV_4151" in C:\WINXP\Inf\atiixpag.inf; Device: "RADEON 9600 SE Family (Microsoft Corporation)"; Driver: "RADEON 9600 SE Family (Microsoft Corporation)"; Provider: "ATI Technologies Inc."; Mfg: "ATI Technologies Inc."; Section name: "ati2mtag_RV350".
#I023 Actual install section: [ati2mtag_RV350]. Rank: 0x00002001. Effective driver date: 03/23/2004.
#-166 Device install function: DIF_SELECTBESTCOMPATDRV.
#I063 Selected driver installs from section [ati2mtag_RV350] in "c:\winxp\inf\atiixpag.inf".
#I320 Class GUID of device remains: {4D36E968-E325-11CE-BFC1-08002BE10318}.
#I060 Set selected driver.
#I058 Selected best compatible driver.
#-124 Doing copy-only install of "PCI\VEN_1002&DEV_4151&SUBSYS_7C37174B&REV_00\4&38B71F77&0&0008".
#-166 Device install function: DIF_REGISTER_COINSTALLERS.
#I056 Coinstallers registered.
#-166 Device install function: DIF_INSTALLINTERFACES.
#-011 Installing section [ati2mtag_RV350.Interfaces] from "c:\winxp\inf\atiixpag.inf".
#I054 Interfaces installed.
#-166 Device install function: DIF_INSTALLDEVICE.
#I123 Doing full install of "PCI\VEN_1002&DEV_4151&SUBSYS_7C37174B&REV_00\4&38B71F77&0&0008".
#I121 Device install of "PCI\VEN_1002&DEV_4151&SUBSYS_7C37174B&REV_00\4&38B71F77&0&0008" finished successfully.
[2011/11/03 11:05:23 552.1305 Driver Install]

Regards, ska

Last edited by ska (2011-11-05 00:36:50)

Re: [SLVD] G\A\CX_95952.cat is a one-byte-file

Yes, and if you unpack the official AMD/ATI drivers from here you'll see the same thing.
Complain to AMD/ATI for not properly signing their drivers.

Re: [SLVD] G\A\CX_95952.cat is a one-byte-file

mr_smartepants wrote:

Yes, and if you unpack the official AMD/ATI drivers from here you'll see the same thing.
Complain to AMD/ATI for not properly signing their drivers.

Oh, I see, I will complain.

May I suggest to remove any invalid certificates from the DriverPacks, or rename them to XYZ.cat_invalid or something like that? I removed the invalid catalogue file and DPInst could at least install the driver with help of the driver signer warner disabler tool and the Finisher did installed the CCC eventually. The invalid file caused DPInst to log an error, but ignore the driver otherwise.

Thanks mr_smartepants,

ska

Re: [SLVD] G\A\CX_95952.cat is a one-byte-file

You're complaint is valid and noted.  In the future I'll remove .cat files if they interfere with integration.

Re: [SLVD] G\A\CX_95952.cat is a one-byte-file

Just for the archive:

mr_smartepants wrote:

Complain to AMD/ATI for not properly signing their drivers.

My complain at the AMD Global Customer Care resulted into:

If you wish to use WHQL certified drivers for your legacy product, try the Catalyst 9.3 drivers.
Any driver version past this will not be WHQL certified and this will not change.  Drivers past 9.3 for the legacy products are hotfix releases and do not get passed by Microsoft certification, without certification from MS they do not get to be designated as WHQL certified and the code cannot be changed without Microsoft's approval/certification.

This reply does not answer my question about the invalid cat file, though.

Re: [SLVD] G\A\CX_95952.cat is a one-byte-file

That sounds like a BS answer to me.  They even WHQL their beta drivers so I don't understand their answer either.

Re: [SLVD] G\A\CX_95952.cat is a one-byte-file

that 9.3 driver bit rang a bell to me.... found this

AMD may periodically provide Windows XP and Windows Vista driver updates (for the products listed above) for critical fixes only.  No new features will be provided in future driver updates.

All future ATI Catalyst™ releases made available past the ATI Catalyst™ 9.3 release will not include support for the legacy products listed above or any of the features associated with those legacy products.

http://support.amd.com/us/gpudownload/w … ng=English

Last edited by botus (2011-11-25 11:35:42)

Re: [SLVD] G\A\CX_95952.cat is a one-byte-file

So, that begs the question:
-Do we roll back to Catalyst 9.3(.1) to maintain signing, or press on with the unsigned 10.2 version that shows as current on ATI's website?

Re: [SLVD] G\A\CX_95952.cat is a one-byte-file

On the link I gave, it has release notes for 9.3 and yet the recco driver now is 10.2

Which if any of the text they provided there makes sense. 10.2 contained security / critical fixes for devices that functionally gain nothing after 9.3 which makes 10.2 the better choice.

Which if we now go back to what ska had back from AMD that matches...  and 10.2 has bits u need but you can't get it with WHQL

Last edited by botus (2011-11-26 04:03:34)

Re: [SLVD] G\A\CX_95952.cat is a one-byte-file

How about one pack WHQL-only, one pack "3rd-party" (& one 'BASE' to rule them all!)  This could enhance source-security.

  I think a '3rd-party' identifier in the filename could be helpful.  ie: 3P_Graphics_D_wnt5-x86_1111.7z

About the lack of driver signing in legacy AMD drivers:  Contact the new CEO, Rory Read; just be tactful & patient.  He may have the clout to appease owners of legacy hardware.  I can 'see' the headlines already "AMD extends support for Windows 8 to legacy hardware owners",  "AMD says 'buy a new one, when YOU are ready', with enhanced trade-in offers included in advertising with new drivers".  What a contrast this would be to nV & Chipzilla!

http://www.facebook.com/rpread

Last edited by TechDud (2011-11-28 03:19:47)

11

Re: [SLVD] G\A\CX_95952.cat is a one-byte-file

Finally, the reply regarding the one-byte .cat file:

If it is an error it will be corrected, the driver developers have been forwarded this feedback as well.  If it is the way it is supposed to be, it will remain as is.

Re: [SLVD] G\A\CX_95952.cat is a one-byte-file

@ska, thank you for your persistence!  Hopefully they'll listen to reason and do the right thing.