Date

Driver Revision

             Description

Build Number

July 29, 2004

Beta 14.7.0.3877

Weekly engineering drop; not fully validated

3877

July 22, 2004

Beta 14.7.0.3874

Weekly engineering drop; not fully validated

3874

July 15, 2004

Beta 14.7.0.3873

Weekly engineering drop; not fully validated

3873

July 8, 2004

Beta 14.7.0.3868

Weekly engineering drop; not fully validated

3868

July 2, 2004

Beta 14.7.0.3863

Weekly engineering drop; not fully validated

3863

June 24, 2004

Beta 14.7.0.3859

Weekly engineering drop; not fully validated

3859

Issues Resolved in 14.7.0.3874

Reference No.

Description

Affected Component(s)

Affected OS(s)

Affected Project(s)

BugID: 1141661

Associated Info:

SightingsID: 45576

[Clone mode]Overlay of CRT may not be displayed at the time of DVD play back.

GDI

Windows* XP

Intel(r) 855GM Chipset 

Resolution Description:

Details: When we go to DDC from extended desktop (with pasted excel sheet), DrvDisableSurface( ) is not getting called (looks like since pasted excel object is associated with extended desktop’s PDEV, DrvDisableSurface( ) is not getting called to free the PDEVs associated with extended desktop and just DrvAssertMode ( ) is getting called to disable these PDEVs). At this point of time DISPLAY_CONFIGURATION_DPSIM flag is set properly. Now when we kill the Excel in DDC, DrvDisableSurface( ) is called to kill the PDEVs associated with Extended Desktop. Here we don’t check if we have already changed the configuration to DPSIM and hence OR the flag DISPLAY_CONFIGURATION_SPSIM according to the logic. (File: enable.c, line 1029) Now we open WinDVD to play movie, UpdateOverlay32( ) is called. In Updateoverlay32, first driver checks for Single Display flag (SPSIM) and since the flag is True, driver assumes that it is in Single display mode and shows the overlay in one pipe only The Fix Given: The fix is to check DPSIM flag before changing the dwDisplayConfiguration in DrvDisableSurface( ) as this is only flag of which we are not taking care of.

BugID: 1161613

Associated Info:

BugID: 1162354

SightingsID: 47546, 48536

CUI shows primary as active when in DDC mode

GDI

Windows* XP

Intel(r) 855GM Chipset, Intel(r) 855GME Chipset 

Resolution Description:

Details: The issue is that if we open a DD application in Extended desktop and switch to Clone mode, kill the DD application and then check CUI, it shows the primary device as active device though we are in clone mode. Root Cause: The problem is that when we switch to clone mode from extended desktop the call to DrvDisableSurface() is deferred as the DD application is still active. When we kill the application, the deferred call to DrvDisableSurface() is executed and at this point it thinks that we are still in extended desktop and leaving extended desktop and decrements the device count and sets the configuration of SPSD. That’s why we see only the primary active in CUI. The FixGiven: At this point we have the information that we have already switched to clone in one of the deviceconfigurations. The dwdisplayconfiguration of device1 or device2 will be set to DISPLAY_CONFIGURATION_DPSIM. So, before decrementing the device count, check display configuration of the devices. If it is set to DISPLAY_CONFIGURATION_DPSIM, then it indicates that we have already switched to clone. So, do not decrement the device count, because we are still in MDS. Now, since there are two devices active, the display mode is not set to SINGLE_PIPE_SINGLE_DISPLAY_MODE.

BugID: 1372942

Associated Info:

BugID: 1385194, 1387034, 1387308, 1397109, 1401738

SightingsID: 61074

Switching to Dual Display Clone from Extended Desktop through CUI not possible after running a 3D screensaver

GDI

Windows* XP

Intel(r) 915G Chipset 

Resolution Description:

Root Cause: When we run a 3D app, call to DrvDisableSurface is delayed because current instance of the driver owns a Direct3D, WNDOBJ, or DRIVEROBJ. As a result of this delay bMDSEnable flag is still TRUE in HwEnable function. Secondly the fix for 1161613 was causing the ulDev0PDEVCount & ulDev1PDEVCount to go out of synch as it was decremented conditionally and incremented everytime. The Fix Given: 1. Alternative fix has been put for 1161613 in a way as not to disturb the DEVCount’s. 2. Instead of putting an assert and setting dwDisplayConfigurationFlags to DISPLAY_CONFIGURATION_MDS, bMDSEnable is set to false to indicate we are not in MDS mode any more.

BugID: 1376566

Associated Info:

SightingsID: 60856

Primary display on Tablet

GDI

Tablet PC*

Intel(r) 855GM Chipset 

Resolution Description:

same source fix as under DCN 318520

BugID: 1385194

Associated Info:

BugID: 1372942, 1387034, 1397109

SightingsID: 62686

DVI is not available in CUI when using user fast switch after S4 and changing view setting

GDI

Windows* XP

Alviso 

Resolution Description:

Same source fix as 1372942.

BugID: 1387034

Associated Info:

BugID: 1372942, 1385194, 1397109

SightingsID: 62686, 62748

CUI cannot re-open when using user fast switch back to primary user after screensaver and changing view setting

GDI

Windows* XP

Alviso 

Resolution Description:

Same source fix as 1372942.

BugID: 1387308

Associated Info:

BugID: 1372942

SightingsID: 62342

Cannot change to Clone mode when play a DVD movie during enable extend desktop mode.

GDI

Windows* XP

Intel(r) 855GM Chipset 

Resolution Description:

same source fix as 1372942

BugID: 1397109

Associated Info:

BugID: 1372942, 1385194, 1387034, 1401738

SightingsID: 60786

LCD panel is no longer available when applying S3/resume using the lid switch.

GDI

Windows* XP

Alviso 

Resolution Description:

BugID: 1398477

Associated Info:

SightingsID: 63490, 63560, 64424, 64952

WXP2K:System hang after error string displayed while DVD playing on battery

ACPI/Power Management

Windows* 2000, Windows* XP

Alviso 

Resolution Description:

Workaround is to disable the status page feature till the hardware issue is fixed. This would not affect the behavior of DFGT as it is a optional feature.

BugID: 1399213

Associated Info:

BugID: 1426871

SightingsID: 64584

Resolution/color are changed when returning to the Windows screen after using AP that changes the resolution/color.

CUI

Windows* XP

Alviso, Intel(r) 855GM Chipset, Intel(r) 915G Chipset 

Resolution Description:

BugID: 1401738

Associated Info:

BugID: 1372942, 1397109

SightingsID: 64967

LCD is unrecoverable with neither a reboot nor shut down (see long descripton)

GDI

Windows* XP

Alviso 

Resolution Description:

Same source fix as 1372942

BugID: 1403478

Associated Info:

SightingsID: 64510

Auto Dim mode doesn't work in S3 resume state.

Resource Manager

Windows* XP

Alviso 

Resolution Description:

Root Cause: Restoring stale LBPC register contents in the Graphics Driver is causing this problem during a resume event. Fix Description: In order to fix the problem the following steps have been taken: 1. Driver will restore the LBPCvalue after a resume event only if the current content of the LBPC register is the power on default (0x00h) 2. BIOS will ensure that always a NON-ZERO value will be set for the backlight.

BugID: 1404996

Associated Info:

SightingsID: 65317

[SDG+TV] BSOD when TV enabled with MCCS application installed on system.

Soft BIOS

Windows* XP

Intel(r) 865G Chipset 

Resolution Description:

Root Cause: MCCS app initiates the MCCS transaction on TV and GPIO_Pin_List get selected to none (zero) and its leads to wrong calculation of pin pair values. Fix Description: invalid pin-pair value will get the out of bound value from GPIOPinMMIOTbl[] table and its trying to write at an illegal memory location and its going to BSOD.

BugID: 1426646

Associated Info:

SightingsID: 65491

Corruption / DVD CyberLink PowerDVD 5, HCT11.2, Janet Jackson DVD movie - Corruption images are displayed on PowerDVD 5 playback window while running the HCT 11.2 WHQL test kit [REG]

DVD

Windows* XP

Intel(r) 915G Chipset 

Resolution Description:

Root Cause: This problem happens in PowerDVD only, not in WinDVD, MCE, and WMP. DdGetMoCompBuffInfo was called multiple times; in the first 3 times, the width/height passed in were 720/480; in the later 2 times, they are 720/576. We always reported the number of macroblocks based on the width/height passed in. However, the app kept to use the one based on 720/480. It caused DdRenderMoComp called twice per frame. Fix Description: It is a temp fix and will be removed after CyberLink fixes it in the app. Force to use 576 instead of 480 when we receive the width as 720.

Reference No.

Description

Affected Component(s)

Affected OS(s)

Affected Project(s)

BugID: 1141661

Associated Info:

SightingsID: 45576

[Clone mode]Overlay of CRT may not be displayed at the time of DVD play back.

GDI

Windows* XP

Intel(r) 855GM Chipset 

Resolution Description:

Details: When we go to DDC from extended desktop (with pasted excel sheet), DrvDisableSurface( ) is not getting called (looks like since pasted excel object is associated with extended desktop’s PDEV, DrvDisableSurface( ) is not getting called to free the PDEVs associated with extended desktop and just DrvAssertMode ( ) is getting called to disable these PDEVs). At this point of time DISPLAY_CONFIGURATION_DPSIM flag is set properly. Now when we kill the Excel in DDC, DrvDisableSurface( ) is called to kill the PDEVs associated with Extended Desktop. Here we don’t check if we have already changed the configuration to DPSIM and hence OR the flag DISPLAY_CONFIGURATION_SPSIM according to the logic. (File: enable.c, line 1029) Now we open WinDVD to play movie, UpdateOverlay32( ) is called. In Updateoverlay32, first driver checks for Single Display flag (SPSIM) and since the flag is True, driver assumes that it is in Single display mode and shows the overlay in one pipe only The Fix Given: The fix is to check DPSIM flag before changing the dwDisplayConfiguration in DrvDisableSurface( ) as this is only flag of which we are not taking care of.

Issues Resolved in 14.7.0.3873

Reference No.

Description

Affected Component(s)

Affected OS(s)

Affected Project(s)

BugID: 1390335

Associated Info:

SightingsID: 63559

When play game "Halo" The display had disorderly

D3D

Windows* XP

Intel(r) 855GM Chipset 

Resolution Description:

It was found that having the maxstreams capability set to 16 was causing texture corruption on pre-Napa family platforms (MGM, SDG, etc). A code adjustment has been made to set the value to 1 for all pre-Napa platforms and 16 for Napa platforms.

BugID: 1396916

Associated Info:

SightingsID: 64403

DFP hotpugging/unplugging make display come in TV only mode

Soft BIOS

Windows* XP

Intel(r) 865G Chipset 

Resolution Description:

Description and Fix of the bug : When we hot Unplug the DFP and switch to TVto Di, Settiming sable the Port and the DFP encoder was not succeeding. This is the reason we see display on DFP. We have a fix in settiming which does not allow to Disable/Enable or program the encoder if DFP is not connected. This fix holds good if we are trying to enable or program the encoder with a new timing. For Disabling the encoder we need not have DFP connected.

Issues Resolved in 14.7.0.3868

Reference No.

Description

Affected Component(s)

Affected OS(s)

Affected Project(s)

N/A

N/A

N/A

N/A

N/A

Issues Resolved in 14.7.0.3863

Reference No.

Description

Affected Component(s)

Affected OS(s)

Affected Project(s)

BugID: 1385371

Associated Info:

SightingsID: 63063, 63871, 64055

With only TV connected, CRT device is showing in the CUI devices tab

Soft BIOS

Windows* XP

Alviso 

Resolution Description:

Root Cause and Fix : CRT Detection was failing on certain systems…..specially Alviso SouthBay systems. We need to give a small delay around 1Micro seconds after programming the registers. Made sure that this will not cause any flicker on DFP,CRT and LFP.

BugID: 1396799

Associated Info:

SightingsID: 64302

Position - Flicker screen is displayed while clicking position button to move TV screen position

AIM Module

Windows* XP

Intel(r) 855GM Chipset 

Resolution Description:

Bug Title ========= 1396799 – Flicker screen is displayed while clicking position button to move TV screen. 1398657 – Focus 454 issues. (TV flickers while moving horizontal & vertical sliders and also corruption issue) Details ======= These two Tibet bugs report bunch of issues related to FS454 card. These issues are: 1)Flicker on the screen while clicking the position buttons to move TV screen. 2)Flicker on the screen while moving horizontal & vertical sliders 3)Display corruption when horizontal slider is at 1000 and vertical slider is around 124. The root cause for (1) & (2) is Focus driver does full programming whenever there is change in X/Y position and also during overscan (horizontal & vertical sliders) changes. The full programming sequence turns off DACs, programs all registers and turns on DACs. This sequence results in screen flicker due to turn off and turn on. The changes in following files attempt to resolve the above issues in the following manner. 1)Do not do full register programming for X/Y changes. With this change, no flickers observed when changing X positions. However some amount of flicker is observed when changing Y position, this seems to be hardware limitation. 2)Flicker due to overscan changes seems to be unavoidable due to full programming. One may argue that there are less flickers observed in Ch7009. Ch7009 has two or three overscan ranges and whenever cross over of range happens, flicker is observed. However in case of focus, such range does not exist and for every overscan change, flicker is seen. 3)Display corruption is fixed by tying both horizontal & vertical overscan values. In other words, whenever dwOverscanX changes, change dwOverscanY value by the same magnitude and program them together. The end behavior is much like ch7009. Files Changed ============= 1)$gfx_Development\Source\aim\A304\fs453.c 2)$gfx_Development\Source\aim\A304\fs450Dev.c Note: This code checkin has fix for Tibet 1396643

Issues Resolved in 14.7.0.3859

Reference No.

Description

Affected Component(s)

Affected OS(s)

Affected Project(s)

BugID: 1386132

Associated Info:

SightingsID: 62490, 62836

CLONE: GMCH CV PERF GDG: Webmark2004 failing intermittently running offline mode benchmark on IGD

D3D

Windows* XP

Intel(r) 915G Chipset 

Resolution Description:

Reviewing this issue with D3D development it was determined that the failing issues all reported O/S API calls. The First one is a DR Watson it reports Iexplorer generated an zappkicaotn error. at address 5c00d80f(D3DIM700Destroy Texture), this occurs along with the following errors in event viewer Faulting module DDraw.dll and faultying module d3dim700.dll The GPF output reports 0xc0000005 - Null pointer. whiel it is noted that the previous platofmrs do not exhibit this failure it is confimred that multiple textures that were supported previously have been removed. to confimr this is turely a applicaiotn error we will need to work with the vendor and have them review there applicaiotn source code It has not been sucessfully determined that the graphic driver is not at fault in the Webmark 04 failures. When disabling DX acceleration, DX calls no longer get to the driver so the application completes without erroring out. This shows that there may be some dependencies to the graphic driver. It is agreed that the application vendor source code needs to be reviewed and there may be possible problems due to this benchmark coming out last year prior to GDG graphics. *** wbdavids 6/21/2004 *** A driver change was made to change the cap we report to DX7 and earlier applications for wMaxSimultaneousSurfaces from 16 to 8. We should not be reporting more support than the API can support because some applications key off of that capability instead of the limit of the API. This was also poorly speced by Microsoft. Initially they wanted us to report the number of samplers in this cap regardless of API capabilities. They have since changed their minds.

BugID: 1387576

Associated Info:

SightingsID: 63430

Persistence - display position is moved to center after clicking Up position button when setting display position to maximum left.

Soft BIOS

Windows* 2000, Windows* XP

Alviso 

Resolution Description:

Same Source Fix as 1395086

BugID: 1394910

Associated Info:

SightingsID: 63531

TV display becomes very dark when attempting to change TV format

TVout

Windows* XP

Alviso 

Resolution Description:

Same Source Fix as 1395086

BugID: 1395086

Associated Info:

BugID: 1374227

SightingsID: 61387, 64530

TV settings: Flicker settings adjustment values are opposite as compared with MGM

TVout

Windows* XP

Alviso 

Resolution Description:

I have included 3 minor TVOUT issues(1394910,1395086,1387576) into one. These are the following issues and their details. 1394910- TV display becomes very dark when attempting to change TV format Root Cause : TV parameters from the VIDEOPARAMETER structure in registry was being applied without doing a check to see if the corresponding flag for the parameter is set. 1395086 TV settings: Flicker settings adjustment values are opposite as compared with MGM Root Cause and Fix : The flicker filter itself is working correctly. The slide bar values used to define the flicker filter needs to be reversed to keep consistency across platforms. Thus, a value of 0 would provide no filter (flickering the most visible) and a value of 1000 would provide the maximum filetering capability (flickering reduced to optimum capability of HW) 1387576 [VAL-AIM] 915GM: Persistence - display position is moved to center after clicking Up position button when setting display position to maximum left. [RTA][REG] Root Cause and Fix: Currently we were setting the position of the display to 50% when xPos and yPos are equal to zero. This was based on the Ch7009 code. This is not required. We need to set proper user selected values.

BugID: 1395351

Associated Info:

SightingsID: 63824, 64486, 64682

Extended desktop configuration makes system hang

GDI

Windows* XP

Alviso, Intel(r) 855GME Chipset 

Resolution Description:

Fixed by: hjayanna Cause: Display Change event was getting triggered well before SharedInfo data structure was completely initialized. This was causing an access violation in the Ring Programming code during GSV frequency switch. Secondly during the Display Change event the HW_DEVICEEXTENSION pointer is updated in the PC_CONTEXT, this was causing the BIA timer routine to update the palette for an incorrect pipe. Fix: Trigger Pre-Mode Set event from GDI in HwAssertMode() call just before the mode is set. For the second problem the solution is to update the Gamma only when PIPE_B is the current pipe index in the DEVICE_EXTENSION.