DirectDraw Test

DirectDraw contains a series of tests that perform basic checks on a Display Driver's DirectDraw features.

Algorithm

The DirectDraw test performs the following specific tests:

1. PC98_22 (PC97 20) Perform Double Buffer Swaps with no tearing
   Tests a drivers support for page flipping. Tries to verify whether or not page flipping is in sync with the vertical blank period. This test is somewhat timing dependent, and should not be run with with a minimal amount of background processes, and with all debug spew turned off.
2. PC99 14.8 ScrnRes & VidMem Amt
   This test verifies that the display drivers meet basic screen resolution and Video Memory PC99 requirements (14.8). PC99 requires 4 MB of Local Video Memory and support for the following screen resolutions:
   • 640x480 [8, 16, 24(or 32) bpp]
   • 800x600 [8, 16, 24(or 32) bpp]
   • 1024x768 [8, 16 bpp]

   The following resolutions are recommended:

   • 1024x768 [24(or 32) bpp]
   • 1280x1024 [8, 16, 24(or 32) bpp]
3. PC99 14.23 Low-Res Support
   This test verifies that the display drivers/ Hardware support linear-mapped, low-resolution modes (14.8). The following screen resolutions are required:
   • 320x200 [8, 16 bpp]
   • 320x240 [8, 16 bpp]
   • 640x400 [8, 16 bpp]

   The following resolutions are recommended:

   • 400x300 [8, 16 bpp]
   • 512x384 [8, 16 bpp]
4. PC99TA 6.8.1 Overlay support
   This test checks for basic overlay support.
5. PC99TA 6.8.2 Color key support
   This test checks for overlay support with destination color keying.
6. PAL_CreatePalette - Init 256 clrs(TC 2.6.9)
   This test simply displays a range of colors in various modes. In 8-bit modes a labled palette is displayed.
7. MEMMAN - Consume All Video Mem Test(TC 6.1.1)
   This test allocates surfaces of smaller and smaller sizes such that every usable byte of video memory is consumed (including any AGP memory). As each surface is created, it's buffer is filled with a somewhat unique byte value, which can then be checked upon release to verify the surface integrity and ensure no surface allocations overlapped. The test will allow for 1% (or at least 2K) of video memory to be inaccessible.
8. MEMMAN - Rand Create/Release Video Memory surfaces 100X Test(TC 6.1.3)
   This test makes 100 random calls to create or release a video memory surface. Each created surface is filled with the byte value representing it's number. For example, the first surface is filled with '1'. Upon release, the entire surface is scanned to verify it's contents have not been changed. Created surfaces randomly range in size up to 320x175.
9. MEMMAN_GetAvailableVidMem(TC 3.6.?]
   This test reports the amount of total local and AGP video memory, and performs some basic checks on the numbers.
10. BLT Basic, VMem to Primary(TC 3.6.48, 3.6.103)
    This test performs a basic copy blt from an offscreen video memory surface to the primary. The blt is simulated using generic software blitters for comparison. Both the simulation, and an exhanced XOR of the differences are shown on screen to aid in debugging.
11. BLT, single SRCKEY, VMem to Primary(TC 3.6.60)
    This test performs a source color key blt from an offscreen video memory surface to the primary. The blt is simulated using generic software blitters for comparison. Both the simulation, and an exhanced XOR of the differences are shown on screen to aid in debugging.
12. BLT Color Fill Primary(TC 3.6.57)
    This test performs a series of color fill blts on the primary surface (the screen). The blt is simulated with GDI operations on an offscreen bitmap for comparison (not shown).
13. BLT, Clipped, VMem to Primary Test
    This test performs a series of blts on the primary surface (the screen) where an elliptical clipping region has been established. The blt is simulated on an offscreen bitmap for comparison (not shown).
14. BLT_OnscreenOverlap
    This test attempts to perform overlapping blts using the primary surface as both the source and destination. The blts are performed in a circular pattern and simulated with software blitters for comparison.
15. SURF_LockRectVideo
    Simple test which verifies that the appropriate memory address is returned when locking the primary.
16. BLT Stretch x 2, VMem to Primary(TC 3.6.49-50)
    This test performs a stretch blt from an offscreen video memory surface to the primary. The blt is simulated using generic software blitters for comparison. Both the simulation, and an exhanced XOR of the differences are shown on screen to aid in debugging.

Configuration

A DirectX 6 capable video adapter and driver is required.

Issues

The Windows taskbar may cause this test to fail if it appears while this test is running. To avoid possible failures of this test, disable the "Always on top" feature of the Windows taskbar before running this test:

To disable the "Always on top" feature of the Windows taskbar:

1. Right-click on the right side of the Windows taskbar (over the clock).
2. On the drop-down menu, select Properties.
   The Taskbar Properties dialog appears.
3. On the Taskbar Options tab of the Taskbar Properties dialog, clear the "Always on top" check box (i.e. ensure that it is not checked).
4. Click OK.

DPMS Test

The test issues DPMS commands to the display adapter, testing power levels as per the DPMS specification. The screen will blank and flicker momentarily as the test runs.

Parameters

None.

Driver Scenario Test

This test tests that the video adapter/driver combination correctly supports the features that it claims to support. Specifically, the following features are tested:

• Specular highlighting
• Fog
• Alpha blending
• Texturing
• Texture Filtering
• Mip-mapping

The test will run all combinations of the required features in a dynamic scene. The scene is composed of a central, unmoving sphere and six orbiting cubes. Additionally there is a plane extending from the point of view (camera) into the distance, demonstrating correct texture perspective. The cubes and central sphere will take on various properties as the test proceeds, cycling through the supported feature set.

Configuration

A DirectX 6 capable video adapter and driver is required. This video adapter must be set to 800x600, 16 color, resolution before running this test, as specified in the PC 99 System Design Guide. The video mode must have a refresh rate of at least 75hz as measured by the test. In choosing texture formats, this test will only select either a 1:5:5:5 format or a 4:4:4:4 format. Support for one of these formats is required by section 14.30 of the PC 99 System Design Guide, so at least one of these formats must be supported.

Issues

The Windows taskbar may cause this test to fail if it appears while this test is running. To avoid possible failures of this test, disable the "Always on top" feature of the Windows taskbar before running this test:

To disable the "Always on top" feature of the Windows taskbar:

1. Right-click on the right side of the Windows taskbar (over the clock).
2. On the drop-down menu, select Properties.
   The Taskbar Properties dialog appears.
3. On the Taskbar Options tab of the Taskbar Properties dialog, clear the "Always on top" check box (i.e. ensure that it is not checked).
4. Click OK.

Special Considerations for Mobile Systems

This test's 75hz refresh rate display requirement is not possible with the displays included on many mobile systems. Moblie systems that lack 75hz display capabilities may be tested using an external monitor, if external monitors are supported. Mobile systems with displays that are not capable of a 75hz refresh rate and that do not support external monitors will not be required to pass this test.

Driver Scenario Test (PC 99)

This test tests that the video adapter/driver combination correctly supports the features that are required by the PC 99 System Design Guide. Specifically, the following features are tested:

• Specular highlighting
• Fog
• Alpha blending
• Texturing
• Texture Filtering
• Mip-mapping

The test will run all combinations of the required features in a dynamic scene. The scene is composed of a central, unmoving sphere and six orbiting cubes. Additionally there is a plane extending from the point of view (camera) into the distance, demonstrating correct texture perspective. The cubes and central sphere will take on various properties as the test proceeds, cycling through the required feature set.

Configuration

A DirectX 6 capable video adapter and driver is required. This video adapter must be set to 800x600, 16 color, resolution before running this test, as specified in the PC 99 System Design Guide. The video mode must have a refresh rate of at least 75hz (60hz cap for laptops??) as measured by the test. In choosing texture formats, this test will only select either a 1:5:5:5 format or a 4:4:4:4 format. Support for one of these formats is required by section 14.30 of the PC 99 System Design Guide, so at least one of these formats must be supported.

Issues

The Windows taskbar may cause this test to fail if it appears while this test is running. To avoid possible failures of this test, disable the "Always on top" feature of the Windows taskbar before running this test:

To disable the "Always on top" feature of the Windows taskbar:

1. Right-click on the right side of the Windows taskbar (over the clock).
2. On the drop-down menu, select Properties.
   The Taskbar Properties dialog appears.
3. On the Taskbar Options tab of the Taskbar Properties dialog, clear the "Always on top" check box (i.e. ensure that it is not checked).
4. Click OK.

Special Considerations for Mobile Systems

This test's 75hz refresh rate display requirement is not possible with the displays included on many mobile systems. Moblie systems that lack 75hz display capabilities may be tested using an external monitor, if external monitors are supported. Mobile systems with displays that are not capable of a 75hz refresh rate and that do not support external monitors will not be required to pass this test.

IDE and Video Disable

This test is designed to verify that the IRQs normally reserved for the floppy disk drive controller (IRQ 6) as well as the IDE hard disk controller (IRQs 14 and 15) can be disabled and are usable by other devices. If your system does not use IDE devices and another device is already using the IRQ normally reserved for IDE devices, then the IRQ is considered usable and further IRQ testing of it may be skipped. The Onboard program can be found in the TESTBIN directory on the HCT CD.

Testing Procedure

1. Power on the system and enter the BIOS setup screen.
2. Disable the on-board IDE controllers as well as the floppy disk drive. Set any IDE device and floppy device values to None or Disabled. Set the BIOS to PnP OS or allow the OS to control IRQ's. Save the new settings and exit the BIOS.
3. Power off the system and install a PCI SCSI hard disk controller and SCSI hard disk. Power on the system and proceed to install Windows 98 on the SCSI hard disk.
4. After completion of the new Windows 98 installation, open the "System" Control Panel applet, and click the "Device Manager" tab. Double click on the "Computer" from the root of the device listing. If any device is using either IRQ 6, 14 or 15, then that used IRQ is considered disabled and free for use by other devices and further IRQ testing of it may be skipped.
5. Now make note of the IRQ's that are not in use. Next, click the "Reserve Resources" tab and "Add…" all IRQ's that are not currently in use except for 6, 14 and 15. At this point, you will have effectively used up all free IRQ's by other devices or the OS itself.
6. Reboot the system and load Windows 98. Again, enter the "System" Control Panel applet, click the "Device Manager" tab, double click on the "Computer." Look at the IRQ's in use. Verify that other devices are using IRQ's 6, 14 and 15. If any of the three IRQ's (6, 14, or 15) are not in use, click the "Reserve Resources" tab and "Add…" IRQ's that other PCI devices may be using. This will force other devices to migrate over to the IRQ's to be tested. Repeat this procedure after each reboot until you have successfully verified that other devices have used each of the three IRQ's.
7. Run the Onboard program and verify that the I/O ranges are free for the floppy controller and the IDE controllers.
8. After you have succeeded, be sure to remove the IRQ's from the "Reserved Resources" so that Windows can restore the reserved IRQ's to other devices. Now power down the system and remove the SCSI controller and SCSI hard drive.
9. Power the system back up, enter the BIOS setup screen, and restore all IDE and floppy device settings to their initial state.

Video Chipset Disable Instructions

1. Insert a PCI video card using the same chipset (if possible).
2. Run AVI Playback test with add in card to verify there is no corruption in any of the video modes.

Note: it is recommended you test many different add-in cards from different manufactures, one manufacture's card may function fine while another manufacture's card will hang the system.

Recommended chipsets: S3, ATI, Cirrus Logic, Matrox

Video (Win GDI w/ poly. clip_256)

This test cycles through a series of metafile variations contained in a script file and generates them on both a video device context (DC) and a memory DC. The test then compares the two DC surfaces for differences. Polygonal clipping functions are also used. The test logs the results of these comparisons and writes a summary to the log file.

Parameters

None.

Issues

Do not run any programs in the foreground of the test application window or change the focus because this can invalidate test results. Moving the mouse during this test will also cause iterations to fail, and network broadcast messages can invalidate the test results.

Video Playback

This test plays an AVI file (windowed and full-screen) in all supported color depths and resolutions. After each resolution has been tested, the user will be asked if the AVI file was displayed twice and if sound was heard while the AVI file was displayed. Answers to this prompt are recorded in the test log. In addition, this test polls the video driver for its capabilities, verifying that it is a true 32-bit driver.

Parameters

None.