|
|
|
|
|
                   |
|
| Release | Date |
Changes |
| 5-00-02 BETA |
16/5/95 |
The PCI IDE timings have been corrected to prevent problems encountered with floppy drive initialisation. |
| 5-00-05 BETA |
7/6/95 |
The IBM Surepath BIOS version number has
been corrected to 1.3, rather than 0.02. A basic version of the AIS manager has been included to allow certain diagnostic programs to identify the motherboard type. IPR 1158 - The PCI video controller was not being programmed correctly resulting in Windows failing to work with more than 16 colour video modes. This has now been resolved. IPR 1209 - Standard HDU CHS was being used all the time, resulting in drives greater than 504Mb not being used at full capacity. This has now been resolved. IPR 1150 - SIMM sizing was incorrect for certain SIMM combinations. This has now been resolved. Video cards were not being detected, as our on-board video was always 'found' at C000. This has now been resolved. |
| 5-00-06 BETA |
28/6/95 |
Support for mode-2 IDE drives has been
included (SETUP=COMPATIBLE), in addition to high performance mode-3 & mode-4 drives
(SETUP-HIGH PERFORMANCE) Various IPRs concerned the 'loss' of CD-ROM drives following a warm boot. This has now been resolved. ROM 'SETUP' had a menu for 'monitor timings'. This is now un-supported, and the menu has been removed. IPR 1221 - ROM shadowing option for memory regions E8000h - EFFFFh not 'greyed out'. This has been fixed. The 'alarm date' was in American date format. Now resolved. The IBM copyright and version number text was incorrect. This has now been resolved. IPR 1161 - Display font would change during POST. Resolved. If a monochrome video adapter was fitted, a colour video mode would still be selected on exit from POST. This has now been fixed. IDE bus mastering is now enabled. APM functionality is now provided. |
| 5-00-07 BETA |
13/7/95 | IPR1255 Adaptec AHA1542 BIOS did not
work at address C8000. This was caused by the memory region being shadowed by VPD. This
has now been resolved. IPR1242 FDU or HDU failures occurred if power management was enabled. This was caused by the SMI handler corrupting MS-DOS memory regions and has now been resolved. The BIOS signed on as ‘Xen PC’. This has been changed to ‘Apricot BIOS’. ‘SETUP’ has been modified to handle our power management implementation. If CMOS was cleared, ‘SETUP’ would be entered on every boot, reporting a configuration mismatch. This has now been resolved. Monitor blanking did not work. This has now been implemented. HDU power-down did not work. This has now been implemented. The SMI handler was not installed if SMI was disabled. This would cause the ‘Standby’ switch and power management support to not function at all. This has now been resolved. The SMI handler was powering down the CPU after a period of inactivity, even if SMI was disabled. Also, if monitor power management was disabled, the monitor would not be turned off when the ‘Standby’ switch was pressed. These problems have now been resolved. The ‘SETUP’ system summary screen should now report ‘Pentium Overdrive’ when a P55CT is fitted. Please note that this has not been tested. The ‘Standby’ LED would not be lit when in power saving mode. This has now been implemented. |
| 5-00-08 BETA |
17/7/95 | IPR1216 - resolved. The power-on
password will now be requested before 'SETUP' can be entered. IPR1318 - TES140 hangs. Now resolved. |
| 5-00-09 BETA |
31/7/95 | IPR11335 - Machine hangs if memory
sockets MM1 & MM2 are empty, but MM3 & MM4 are occupied. Modified SIMM detection
code resolves this issue. IPR1150 - Certain SIMM combinations were not sized correctly. This has now been fixed. EDO SIMMs are now supported. A potential problem could occur with SMIs being 'missed'. This has been resolved. Default setting for HDU translation was Standard CHS. This has been changed to Extended CHS. The 'Standby' light was on when it should have been off. This has now been resolved. IPR1349 - 'SETUP' always reported the processor speed as 90 MHz. This has now been fixed. In 'Standby' mode the LED was constantly on. It now flashes, to be consistent with the 'home' machine. PnP support is now provided for the NS306 part. The Mitsubishi 3-mode floppy is now supported. The CPU was not being slowed down when either the 'Standby' switch was pressed or the inactivity timer expired. This has now been resolved. |
| 5-00-10 BETA |
10/8/95 | 'SETUP' was being invoked on some
machines with a 262/162 error code - 'Memory parity configuration mismatch'. This was
caused by the parity option being removed from 'SETUP' and hence leaving the CMOS location
undefined, but code remained to actually read this location. This has now been fixed. The base address of the NS306 device has changed from 0x398H to 0x02EH. This has been addressed by BIOS code changes. A potential problem could occur if an SMI happened when reading or writing CMOS - the address offset into CMOS could be lost. This has been resolved. The CPU was not being slowed down when the inactivity timer expired or the 'standby' switch was pressed. This has now been fixed. Potentially, an application or operating system could be 'aware' of the Triton chipset and subsequently alter the SMI handler. The SMM RAM is now totally 'locked' to prevent this happening. The BIOS internal functions 'smi_setInactivity' and 'smi_getInactivity' would hang the machine. This was caused by a coding problem and has been fixed. The BIOS internal function 'smi_SlowCPU' was not slowing down the CPU. This has been resolved. A previous code change resulted in the wrong GPIO bit being set for 3-mode floppy support. This has now been fixed. The 'Prefetch' and 'Posting' enable bits for the secondary IDE interface are now cleared, as these features are not supported by ATAPI devices. This may fix IPR1350 - 'Hard drive disappears on warm boot', but has not been tested. 'Palette snooping' has been added as an option in 'SETUP'. NOTE: Some PnP code has been removed for this release |
| 5-00-11 BETA |
17/8/95 | Diagnostics required a method for
determining whether a machine is a 'normal' PC or 'multimedia ready', i.e. is there a
remote control front panel fitted. A CMOS location has now been defined to provide this
information. PCI interrupts were not being routed. This has now been fixed. The AIS manager was not returning the machine ID. This has now been resolved. The CPU power management SETUP options were not being 'greyed out' when a front panel board was fitted. This has now been implemented. BIOS power management was not working. This has now been fixed. Cards fitted in PCI slot 2 were not having PCI interrupts routed correctly. This has now been fixed. Win95 power management was losing break events. This has now been resolved. PCI bus master cards were experiencing problems on warm boot. This has been fixed. Power management was always being 'greyed out'. This has now been resolved. Unexpected NMIs were occurring during an FERR# (floating point) exception. This has now been fixed. If EMM386 was configured with 'NOEMS' the machine would hang as it went into power save. This has now been resolved. An internal BIOS routine to stop the CPU clock was not working. It has now been fixed. The parallel port was defaulting to ECP. This has now been changed to 'standard'. The second IDE interface was defaulting to 'High performance'. This has now been changed to 'Compatible'. |
| 5-00-12 BETA |
31/8/95 | The standby LED would not flash if the
standby button was pressed but power management was not enabled. The BIOS signon copyright message was '1995-1994'. This has been changed to '1985-1995'. SMIs for the Win95 app. were disabled. This has been fixed. 3-mode floppy support may not have worked correctly. This has been addressed by setting bits 4 and 5 GPIO. For debug, an PCI routines would cause a write to port 0x80. This code has been removed. Disabling the parallel port via PnP services would disable the NS306 completely. This has been resolved. When the CPU had been halted by the Win95 app. it would 'wake' up after a fixed period of time, rather than after a 'break' event. This has been fixed. The RTC alarm was setting the 'periodic' alarm bit. This has been resolved. The RTC alarm has been enabled as a 'break' event. The power management code has been altered to allow correct operation with the Win95 app. and 'break' events. IPRs fixed in this release. IPR1217 - The password prompt would not appear at boot if only an administration password was defined. |
| 5-00-13 BETA |
1/9/95 | IPR 1498 fixed - System fails to warm
boot with no mouse fitted. IDE2 Interface now defaults to ‘Compatible’ instead of ‘High Performance’. Power Management is now disabled correctly. |
| 5-00-14 BETA |
5/9/95 | IPR1554 - BIOS power management does not
work on 75Mhz processor if standby timeout is set to three minutes. This has now been
resolved. PCI palette snooping does not work. Reported by ISEC. This has now been fixed. TES640 was failing on ‘COM2:’ test. This has now been resolved. Various problems existed with ‘stopping’ the CPU for power management. These have been addressed, but a potential problem may now exist with 3-mode floppy support |
| 5-00-15 BETA |
11/9/95 | IPR1569 - RTC alarm was not ‘waking
up’ system from power saving mode. This feature has now been removed. Windows 95 was failing to install due to problems with the PnP code for NS306 UART and LPT ports in protected mode. The code from Ruby has been ported across. If COM2: was disabled in ‘SETUP’ then power management was ‘greyed out’. This has been fixed. |
| 5-00 | 15/9/95 | This release is functionally unchanged from version 5.00.15 BETA. |
| 5-01-01 BETA |
4/10/95 | Windows 95 fix for EMM386.EXE and 16Mb
RAM. Fix for Windows NT and >64Mb RAM. Various power management fixes. |
| 5-01 | 4/10/95 | This release is functionally unchanged from the previous BETA version. |
| 5-02-01 BETA |
10/10/95 | Various problems existed concerning
power management, e.g. when using BIOS power management the system would not always enter
or exit 'deep' green mode using the 'standby' switch and when using the Windows power
management application the system would timeout and enter 'light'/'dark' green mode even
when the system is in use. These problems have now been resolved. 3COM PnP Ethernet cards would try to allocate IRQ 15 to themselves, causing a conflict with the PCI IDE. This has now been resolved by insuring the IRQ 15 does not appear as available to PnP devices. |
| 5-02-02 BETA |
10/10/95 | The previous BETA was built with a rogue Cirrus Logic video BIOS and exhibited screen corruption and problems with Windows 95. This has now been resolved. |
| 5-02 | 13/10/95 | This release is functionally unchanged from version 5-02-02 BETA. |
| 5-03-01 BETA |
18/10/95 | The IR detection code has been temporarily disabled due to intermittent problems with BIOS power management. |
| 5-03-02 BETA |
19/10/95 | A problem existed with the PCI IRQ
allocation (3COM ‘590 Ethernet card). This has now been resolved. If the ‘standby’ button was pressed repeatedly during BIOS POST the machine would hang. This has now been resolved. |
| 5-03-03 BETA |
20/10/95 | If the ‘standby’ button was pressed during BIOS ‘SETUP’ or the ‘virus detection’ screen being displayed, the machine could hang. This has now been resolved. |
| 5-03-04 BETA |
20/10/95 | The monitor standby feature of power
management was not correctly setting the Hsync & Vsync signals to allow the monitor to
enter a low power state. This has now been resolved. After CMOS was cleared and ‘SETUP’ entered, the machine did not recognise the floppy on a subsequent boot. This has now been resolved. |
| 5-03 | 25/10/95 | This release is functionally unchanged from version 5-03-04 BETA. |
| 5-04-01 BETA |
1/11/95 | Support has been re-introduced for
detecting the remote control front panel. The BIOS now reports the date and time of an RTC alarm event to the Windows Power Management application. |
| 5-04-02 BETA |
1/11/95 | This release has been made to correct a build fault which occurred with version 5-04-01 BETA. |
| 5-04-03 BETA |
6/11/95 | When CMOS was cleared, machines would
not boot. This has been fixed. The CMOS default for the remote control was 'Installed'. This has now been changed to 'Not installed'. There is now support for 'Recovery BIOS' for SST flash devices. The Cumana parallel port CD-ROM drive would not work with ECP mode. This has been fixed. The 'Standby' button now works with a front panel fitted. |
| 5-04-04 BETA |
1/12/95 | Melco PRF DIA-0006, parallel port is set
to IRQ 7 in 'SETUP', but actually uses IRQ 5. This was caused by the IRQ not being
properly set during POST and has been resolved. Support for 150 & 166 MHz processors has been included. The power management application timing was inaccurate when using the BIOS power management function to set the Fast Off timer enabling power-down after a period of time. This was caused by integer maths errors in the BIOS and has been resolved. The BIOS recovery code has been altered to support PROM boards containing recovery code. Following a CMOS cleardown, the processor speed was reported as 0MHz following an automatic configuration. This has been resolved. |
| 5-04 | 13/12/95 | This release is functionally unchanged from the previous BETA version. |
| 5-05-01 BETA |
19/12/95 | Support for the Adaptec AHA2940 PCI SCSI
card has been added. Support for the AT&T backplane has been added. |
| 5-05-02 BETA |
5/1/96 | Bus mastering PCI IDE for Windows 95 now
works, with an appropriate driver. Changes have been made for AT&T Dundee, regarding backplane. The remote IR detection code has been altered to address problems encountered with faulty NS306 Silicon (not setting the 'Bad battery bit'). |
| 5-05-03 BETA |
26/1/96 | IPR 1522 - A problem concerning power-on
passwords for user and administrator has been resolved. BIOS recovery code. Fix to power down IDE drives under the Windows 95 power management application. Free up IRQ12 if no mouse is installed. 'Setup' options to disable floppy and both primary and secondary IDE controllers to free up resources if required (Windows 95 requirement). IPR 1224 - PnP memory resource problem. Various changes for AT&T. Support for DIVA card. Melco PRF DIA-0009 - Windows 95 losing 'time' when in power saving mode. |
| 5-05-04 BETA |
2/2/96 | IRQs are now freed up properly. Reported
by AT&T, the PC would hang while initialising the SCSI BIOS if the secondary IDE
controller is disabled. The PnP node information has been corrected for disabled floppy and IDE interfaces. IPR 2014 - ROM 'SETUP' was not being called on F1 if a Diva or Adaptec SCSI card is fitted to the machine. This has now been resolved. |
| 5-05-05 BETA |
8/2/96 | A problem existed when a Diva security
card was installed into a machine which had it's CMOS cleared. A message, 'Enter old SIN
number', would appear. This has now been resolved. Another problem existed when a Diva security card was removed from a system. A message, 'Diva card removed', would appear. The message has now been changed to 'Apricot LocTech 2.1 Security System ...'. Machines would not 'wake up' from standby mode on incoming messages (fax / modem) via COM1: and COM2:. This has now been resolved. ***** 5436 build only ***** A new video BIOS binary from Cirrus Logic has been incorporated. This binary does not display the Cirrus Logic video BIOS signon message. Support for bus mastering IDE drives has been removed. |
| 5-05-06 BETA |
20/3/96 | A fix for PnP option ROM search/execute
has been incorporated for Diva security cards. PRF 00013 PCI IRQ routing for Apricot backplane has been corrected. A change has been made for AT&T to display 'Press a key to continue ...' when a configuration change has been detected. The 'official' Cirrus Logic 5436 binary has been included. ***** This is for 5436 builds only ***** |
| 5-05-07 BETA |
23/2/96 | The setup options to disable floppy and
IDE controllers have been removed. Errors were occurring when trying to execute PCI card option ROMs. This has now been resolved. |
| 5-05 CheckSUM |
29/2/96 F800 |
This release is functionally unchanged from the previous BETA release. |
| 5.06.01 BETA CheckSUM |
19/3/96 A000 |
AT&T request - hard drive CMOS value
setup has been altered. AT&T request - enabling / disabling of secondary IDE controller is now dependant upon whether a device is connected or not. AT&T request - base address of parallel port now cannot be changed. PCI IDE bus mastering is now re-enabled. AT&T request - boot sector virus checking is now disabled regardless of SETUP options. AT&T request - keyboardless mode is enabled regardless of SETUP options. From Windows 95, restarting the system in MS-DOS mode would cause the floppy drive to 'disappear' on some systems. This has now been resolved. |
| 5.06 CheckSUM |
3/4/96 A000 |
This release is functionally unchanged from the previous beta version. |
| 5.07.01 BETA Checksum P/N 154662SW |
26/4/96 D500 |
The object of this release is to provide
the latest software to the field. Diamond BIOS V5.07.01 includes all the mods since V5.05 which were turned off for V5.06 of the BIOS. Therefore the addition since V5.06 are :-
|
| 5.07 checksum PN 154662SW |
17/5/96 C900 |
This release is functionally unchanged from the previous beta version. |
| 5.08.01 BETA checksum PN 154662SW |
10/6/96 3800 |
A problem encountered with an
‘Accel/2’ PCI video graphics adapter has been resolved. A fix for NCR has been provided to address a problem with OS/2 and invalid CMOS contents. A fix for NCR has been provided to address a problem with 8Mb systems. Support for two VGA controllers has been provided for NCR only. There was a problem with PCI video card resources clashing with PCI SCSI card. This has been resolved. |
| 5.08.02 BETA checksum PN 154662SW. |
25/6/96 9300 |
Changes since last release, CMOS defaults have been changed for NCR systems only. Memory configuration has been changed for NCR systems only. ‘Memory hole’ support has been added for NCR VROP video cards fitted as a secondary VGA device. The game port is now disabled for NCR systems. |
| 5.08.04 Beta checksum PN 154662SW. |
17/9/96 FB00 |
Problem : The Diamond Stealth 1000 card
causes the machine to hang after the memory test with a port 80 code 0f 75h. Cause : The stealth card does not seem to be PCI 2.01 complient and is causing a phantom PCI device to appear which contains rubbish and has a BARD containing all F's hence the FMD can't accomodate the required space. Cure : Only scan physically present slots and PCI devices. Problem : The Matrox Millenium card causes the machine to hang after the memory test with a port code of 075h Cause : The PCIVideo routine builds a prefetchable BARD list but does not pre-configure the prefetchable address space, hence when the remaining PCI space is configured it expects to find the space allocated in the FMD list but doesn't and hangs. Cure : As the Diamond chipset does not seem to support the prefetch capability we can ignore this and use the GEN_MEM_BARD_LIST which the PCIVideo does configure. NOTE : Also, Mod 370 saves the context of di as it gets corrupted throughout PCI configuration, there was one invocation that had been missed. This has been corrected in this mod. Problem : Intel 100MB/s PCI Ethernet Card did not function correctly Cause : Memory space was being disabled Cure : Fix from IBM preserved the memory space setting in the command register Note : This fix is on a switch called PCIETHERCARDFIX which is set to TRUE in PCI_POST.ASM. |
| 5.08.05 Beta checksum PN 154662SW. |
21/10/96 8200 |
Changes since last release, PRF 1710 - virus detection code not working. A ‘fix’ for AT&T was implemented incorrectly, thereby disabling this functionality. This has now been resolved. The NCR VEROP card would hang with port 80 code 075. This was caused by a programming error and has now been resolved. PRF1725 - administrator password behaving like the power-on password. This has now been resolved and will only be requested if the F1 setup key is pressed. PRF1614 - LPT1 detected as ECP printer port by Windows 95 when disabled in BIOS setup. This was caused by Windows 95 reading the chipset directly and ignoring CMOS settings. Now resolved. |
| 5.08.06 Beta checksum PN 154662SW |
24/10/96 2C00 |
Changes since last release, PRF 1636 - Adaptec 2940 card and DIVA fail to boot together. This was caused by the DIVA option ROM returning the processor from protected mode to real mode rather than ‘big’ real mode, as required by the PCI configuration code. The result being a general protection fault as the PCI code attempts to examine the FE000000h ROM space - this being above the FFFFh address limit imposed by real mode. This has now been resolved. |
| 5.08 Release checksum PN 154662SW. |
8/11/96 2C00 |
This release is functionally unchanged from the previous BETA version. |
| 5.09.01 Beta checksum PN 154662SW. |
28/11/96 0F00 |
Changes since last release, The GEM line parallel port ‘Genius’ LAN adapter would occasionally fail to connect to the network. This was caused by a coding error in modifications applied for the RTC ‘century roll-over’ issue and has now been resolved. |
| 5.09.02 Beta checksum PN 154662SW |
22/01/97 B200 |
Changes since last release, Booting with Diva card, BIOS revision 5.08.05 and later with 5434 video binary caused a security card change error and the Diva card would request the SIN number of the non-existent ‘old’ card. This was caused by the compresssed video binary over-running the AIS variables and has now been resolved. |
| 5.09.04 Beta checksum PN 154662SW |
10/07/97 7D00 |
Changes since last release, Year 2000 support added. Hard disk drives >4GB support added. |
| 5.09 Release checksum PN 154662SW. |
14/11/97 7D00 |
This release is functionally unchanged from the previous BETA version (5.09.04). |
| Release | Date |
Changes |
| 5-00-15 BETA |
15/9/95 | Support for the 5436 VGA controller part c/w EDO RAM with 80Mhz ‘MCLK’. |
| 5-01-01 BETA |
6/10/95 | The Cirrus Logic 5436 video BIOS was incorrectly setting MCLK to 50Mhz on a warm boot. This has now been fixed. |
| 5-02-01 BETA |
10/10/95 | Various problems existed concerning
power management, e.g. when using BIOS power management the system would not always enter
or exit 'deep' green mode using the 'standby' switch and when using the Windows power
management application the system would timeout and enter 'light'/'dark' green mode even
when the system is in use. These problems have now been resolved. 3COM PnP Ethernet cards would try to allocate IRQ 15 to themselves, causing a conflict with the PCI IDE. This has now been resolved by insuring the IRQ 15 does not appear as available to PnP devices. |
| 5-02 | 13/10/95 | This release is functionally unchanged from the previous BETA version. |
| 5-03-01 BETA |
18/10/95 | The IR detection code has been temporarily disabled due to intermittent problems with BIOS power management. |
| 5-03-02 BETA |
19/10/95 | A problem existed with the PCI IRQ
allocation (3COM ‘590 Ethernet card). This has now been resolved. If the ‘standby’ button was pressed repeatedly during BIOS POST the machine would hang. This has now been resolved. |
| 5-03-03 BETA |
20/10/95 | If the ‘standby’ button was pressed during BIOS ‘SETUP’ or the ‘virus detection’ screen being displayed, the machine could hang. This has now been resolved. |
| 5-03-04 BETA |
20/10/95 | The monitor standby feature of power
management was not correctly setting the Hsync & Vsync signals to allow the monitor to
enter a low power state. This has now been resolved. After CMOS was cleared and ‘SETUP’ entered, the machine did not recognise the floppy on a subsequent boot. This has now been resolved. |
| 5-03 | 25/10/95 | This release is functionally unchanged from version 5-03-04 BETA. |
| 5-04-01 BETA |
1/11/95 | Support has been re-introduced for
detecting the remote control front panel. The BIOS now reports the date and time of an RTC alarm event to the Windows Power Management application. |
| 5-04-02 BETA |
1/11/95 | This release has been made to correct a build fault which occurred with version 5-04-01 BETA. |
| 5-04-03 BETA |
6/11/95 | When CMOS was cleared, machines would
not boot. This has been fixed. The CMOS default for the remote control was 'Installed'. This has now been changed to 'Not installed'. There is now support for 'Recovery BIOS' for SST flash devices. The Cumana parallel port CD-ROM drive would not work with ECP mode. This has been fixed. The 'Standby' button now works with a front panel fitted. A new 5436 video BIOS from Cirrus Logic has been incorporated. This fixes the cold start problem. |
| 5-04-04 BETA |
1/12/95 | Melco PRF DIA-0006, parallel port is set
to IRQ 7 in 'SETUP', but actually uses IRQ 5. This was caused by the IRQ not being
properly set during POST and has been resolved. Support for 150 & 166 MHz processors has been included. The power management application timing was inaccurate when using the BIOS power management function to set the Fast Off timer enabling power-down after a period of time. This was caused by integer maths errors in the BIOS and has been resolved. The BIOS recovery code has been altered to support PROM boards containing recovery code. Following a CMOS cleardown, the processor speed was reported as 0MHz following an automatic configuration. This has been resolved. |
| 5-04 | 13/12/95 | This release is functionally unchanged from the previous BETA version. |
| 5-05-01 BETA |
19/12/95 | Support for the Adaptec AHA2940 PCI SCSI
card has been added. Support for the AT&T backplane has been added. |
| 5-05-02 BETA |
5/1/96 | Bus mastering PCI IDE for Windows 95 now
works, with an appropriate driver. Changes have been made for AT&T Dundee, regarding backplane. The remote IR detection code has been altered to address problems encountered with faulty NS306 Silicon (not setting the 'Bad battery bit'). |
| 5-05-03 BETA |
26/1/96 | IPR 1522 - A problem concerning power-on
passwords for user and administrator has been resolved. BIOS recovery code. Fix to power down IDE drives under the Windows 95 power management application. Free up IRQ12 if no mouse is installed. 'Setup' options to disable floppy and both primary and secondary IDE controllers to free up resources if required (Windows 95 requirement). IPR 1224 - PnP memory resource problem. Various changes for AT&T. Support for DIVA card. Melco PRF DIA-0009 - Windows 95 losing 'time' when in power saving mode. |
| 5-05-04 BETA |
2/2/96 | IRQs are now freed up properly. Reported
by AT&T, the PC would hang while initialising the SCSI BIOS if the secondary IDE
controller is disabled. The PnP node information has been corrected for disabled floppy and IDE interfaces. IPR 2014 - ROM 'SETUP' was not being called on F1 if a Diva or Adaptec SCSI card is fitted to the machine. This has now been resolved. |
| 5-05-05 BETA |
8/2/96 | A problem existed when a Diva security
card was installed into a machine which had it's CMOS cleared. A message, 'Enter old SIN
number', would appear. This has now been resolved. Another problem existed when a Diva security card was removed from a system. A message, 'Diva card removed', would appear. The message has now been changed to 'Apricot LocTech 2.1 Security System ...'. Machines would not 'wake up' from standby mode on incoming messages (fax / modem) via COM1: and COM2:. This has now been resolved. ***** 5436 build only ***** A new video BIOS binary from Cirrus Logic has been incorporated. This binary does not display the Cirrus Logic video BIOS signon message. Support for bus mastering IDE drives has been removed. |
| 5-05-06 BETA |
20/3/96 | A fix for PnP option ROM search/execute
has been incorporated for Diva security cards. PRF 00013 PCI IRQ routing for Apricot backplane has been corrected. A change has been made for AT&T to display 'Press a key to continue ...' when a configuration change has been detected. The 'official' Cirrus Logic 5436 binary has been included. ***** This is for 5436 builds only ***** |
| 5-05-07 BETA |
23/2/96 | The setup options to disable floppy and
IDE controllers have been removed. Errors were occurring when trying to execute PCI card option ROMs. This has now been resolved. |
| 5-05 CheckSUM |
29/2/96 EC00 |
This release is functionally unchanged from the previous BETA release. |
| 5.06.01 BETA CheckSUM |
19/3/96 9400 |
AT&T request - hard drive CMOS value
setup has been altered. AT&T request - enabling / disabling of secondary IDE controller is now dependant upon whether a device is connected or not. AT&T request - base address of parallel port now cannot be changed. PCI IDE bus mastering is now re-enabled. AT&T request - boot sector virus checking is now disabled regardless of SETUP options. AT&T request - keyboardless mode is enabled regardless of SETUP options. From Windows 95, restarting the system in MS-DOS mode would cause the floppy drive to 'disappear' on some systems. This has now been resolved. |
| 5.06 BETA CheckSUM |
3/4/96 8300 |
This release is functionally unchanged from the previous beta version. |
| 5.07.01 BETA Checksum P/N 156009SW |
26/4/96 C600 |
The object of this release is to provide
the latest software to the field. Diamond BIOS V5.07.01 includes all the mods since V5.05 which were turned off for V5.06 of the BIOS. Therefore the addition since V5.06 are :-
|
| 5.07.02 BETA Checksum PN 156009SW |
13/5/96 C000 |
A new Cirrus Logic 5436 video binary has been incorporated. This fixes the issue where the video BIOS was previously writing to an IDE disk controller port. |
| 5.07 checksum PN 156009SW |
17/5/96 B800 |
This release is functionally unchanged from the previous beta version. |
| 5.08.01 BETA checksum PN 156009SW |
10/6/96 2200 |
A problem encountered with an
‘Accel/2’ PCI video graphics adapter has been resolved. A fix for NCR has been provided to address a problem with OS/2 and invalid CMOS contents. A fix for NCR has been provided to address a problem with 8Mb systems. Support for two VGA controllers has been provided for NCR only. There was a problem with PCI video card resources clashing with PCI SCSI card. This has been resolved. |
| 5.08.02 BETA checksum PN 156009SW. |
25/6/96 9300 |
Changes since last release, CMOS defaults have been changed for NCR systems only. Memory configuration has been changed for NCR systems only. ‘Memory hole’ support has been added for NCR VROP video cards fitted as a secondary VGA device. The game port is now disabled for NCR systems. |
| 5.08.03 BETA checksum PN 156009SW. |
20/8/96 5C00 |
Changes since last release Due to incorrect CMOS bank 2 configuration, setting IRQs to ‘not available’ was not completely correct. This has now been resolved. Due to incorrect CMOS bank 2 configuration, PCI error 1801 was being reported when Adaptec 2940 PCI SCSI controllers were present. This has now been resolved. The LogiTech cordless mouse no longer hangs the system during POST. The NCR VEROP (secondary VGA card) now works. After clearing CMOS, IRQ 15 was not available on first boot. This has now been resolved. Floppy drive auto-detection was not completely correct. This has been resolved. The century byte was not being updated on a rollover to the year 2000. This has now been resolved. |
| 5.08.04 Beta checksum PN 156009SW. |
17/9/96 E800 |
Problem : The Diamond Stealth 1000 card
causes the machine to hang after the memory test with a port 80 code 0f 75h. Cause : The stealth card does not seem to be PCI 2.01 complient and is causing a phantom PCI device to appear which contains rubbish and has a BARD containing all F's hence the FMD can't accomodate the required space. Cure : Only scan physically present slots and PCI devices. Problem : The Matrox Millenium card causes the machine to hang after the memory test with a port code of 075h Cause : The PCIVideo routine builds a prefetchable BARD list but does not pre-configure the prefetchable address space, hence when the remaining PCI space is configured it expects to find the space allocated in the FMD list but doesn't and hangs. Cure : As the Diamond chipset does not seem to support the prefetch capability we can ignore this and use the GEN_MEM_BARD_LIST which the PCIVideo does configure. NOTE : Also, Mod 370 saves the context of di as it gets corrupted throughout PCI configuration, there was one invocation that had been missed. This has been corrected in this mod. Problem : Intel 100MB/s PCI Ethernet Card did not function correctly Cause : Memory space was being disabled Cure : Fix from IBM preserved the memory space setting in the command register Note : This fix is on a switch called PCIETHERCARDFIX which is set to TRUE in PCI_POST.ASM. |
| 5.08.05 Beta checksum PN 156009SW. |
21/10/96 6000 |
Changes since last release, PRF 1710 - virus detection code not working. A ‘fix’ for AT&T was implemented incorrectly, thereby disabling this functionality. This has now been resolved. The NCR VEROP card would hang with port 80 code 075. This was caused by a programming error and has now been resolved. PRF1725 - administrator password behaving like the power-on password. This has now been resolved and will only be requested if the F1 setup key is pressed. PRF1614 - LPT1 detected as ECP printer port by Windows 95 when disabled in BIOS setup. This was caused by Windows 95 reading the chipset directly and ignoring CMOS settings. Now resolved. |
| 5.08.06 Beta checksum PN 156009SW. |
24/10/96 1200 |
Changes since last release, PRF 1636 - Adaptec 2940 card and DIVA fail to boot together. This was caused by the DIVA option ROM returning the processor from protected mode to real mode rather than ‘big’ real mode, as required by the PCI configuration code. The result being a general protection fault as the PCI code attempts to examine the FE000000h ROM space - this being above the FFFFh address limit imposed by real mode. This has now been resolved. |
| 5.08 Release checksum PN 156009SW. |
8/11/96 1300 |
This release is functionally unchanged from the previous BETA version. |
| 5.09.01 Beta checksum PN 156009SW. |
28/11/96 F900 |
Changes since last release, The GEM line parallel port ‘Genius’ LAN adapter would occasionally fail to connect to the network. This was caused by a coding error in modifications applied for the RTC ‘century roll-over’ issue and has now been resolved. |
| 5.09.02 Beta checksum PN 156009SW |
22/01/97 9E00 |
Changes since last release, Booting with Diva card, BIOS revision 5.08.05 and later with 5434 video binary caused a security card change error and the Diva card would request the SIN number of the non-existent ‘old’ card. This was caused by the compresssed video binary over-running the AIS variables and has now been resolved. |
| 5.09.04 Beta checksum PN 156009SW |
10/07/97 7700 |
Changes since last release, Year 2000 support added. Hard disk drives >4GB support added. |
| 5.09 Release checksum PN 156009SW. |
14/11/97 7700 |
This release is functionally unchanged from the previous BETA version (5.09.04). |
