[BUG] Solaris x86 2.6 boot hang

Discussion of development and patch submission.
Post Reply
win2kgamer
Posts: 21
Joined: Sun 09 Nov, 2014 12:24 am

[BUG] Solaris x86 2.6 boot hang

Post by win2kgamer » Thu 03 May, 2018 6:41 pm

Solaris x86 version 2.6 hangs while attempting to run the Configuration Assistant with multiple machine types.

This is too early in the boot process to use kadb but here is the message shown when the system hangs:
Solaris26_hang.jpg
Solaris26_hang.jpg (24.83 KiB) Viewed 2611 times
PCem configurations used:
Machine: AMI 486 clone (also tested with the Intel Premiere/PCI and the 430VX)
CPU: Intel i486DX2/66 (as well as a Pentium/66 and a Pentium/90)
Dynamic Recompiler: enabled
Memory: 64MB (also tested with 16MB)
Video: Trident TVGA8900D (as well as a Phoenix Trio64)
Sound: None
HDD: Adaptec AHA-1542C (IO 330, IRQ 11, DMA 5, HBA SCSI ID 7) (also tested with Standard IDE)
FDD1: 3.5" 1.44M
FDD2: 5.25" 1.2M
HDD is on SCSI ID 0 (Primary Master with Standard IDE)
CDROM is on SCSI ID 2 (Secondary Master with Standard IDE)
Mouse: Microsoft 2-button mouse (serial)
Network: None

win2kgamer
Posts: 21
Joined: Sun 09 Nov, 2014 12:24 am

Re: [BUG] Solaris x86 2.6 boot hang

Post by win2kgamer » Fri 26 Apr, 2019 4:05 am

Issue still occurs in latest dev version (04949da), tested with the same configuration as before and with the WinChip interpreter core as well.

win2kgamer
Posts: 21
Joined: Sun 09 Nov, 2014 12:24 am

Re: [BUG] Solaris x86 2.6 boot hang

Post by win2kgamer » Mon 29 Apr, 2019 2:45 am

The boot hang is now fixed as of the latest dev version (0c83cc0). With a Trident TVGA8900, Solaris 2.6 works quite nicely (albeit limited to 256 colors and a little slow).
Screenshot from 2019-04-28 21-42-00.png
Screenshot from 2019-04-28 21-42-00.png (35.81 KiB) Viewed 1871 times

win2kgamer
Posts: 21
Joined: Sun 09 Nov, 2014 12:24 am

Re: [BUG] Solaris x86 2.6 boot hang

Post by win2kgamer » Mon 29 Apr, 2019 2:49 am

While Solaris 2.6 does work with the Trident TVGA8900 (and probably the ET4000 as well), it does reveal a flaw in the S3 graphics emulation:
Screenshot from 2019-04-28 21-04-06.png
Screenshot from 2019-04-28 21-04-06.png (9.74 KiB) Viewed 1869 times

Post Reply