Opened 2 years ago
Last modified 7 months ago
#17886 new bug
[EFI] loader is unable to start the kernel in vmware
Reported by: | diver | Owned by: | |
---|---|---|---|
Priority: | normal | Milestone: | Unscheduled |
Component: | System/Boot Loader/EFI | Version: | R1/Development |
Keywords: | Cc: | davidkaroly | |
Blocked By: | Blocking: | ||
Platform: | All |
Description
This is hrev56385 x86_64 in VMware Workstation 16.
EFI loader displays Haiku splash screen and hangs with no icons lit.
Adding Serial port with output to a text file doesn't produce any output.
Attachments (1)
Change History (9)
comment:1 by , 2 years ago
comment:2 by , 2 years ago
VMWare Worksatation Player 16.2.4 i can reproduce this bug when HDD in Hardware Device list set to IDE and CD/ROM is SATA with connected haiku iso image. When HDD set to SATA or NVMe then Haiku boots successefull
comment:3 by , 2 years ago
Same version here. I tried attaching HDD to SATA and NVME but neither worked for me. Did you leave CDROM attached or are to SATA?
comment:7 by , 2 years ago
I finally managed to boot via EFI after I chose EFI VMware Virtual SATA CDROM from VMware Boot manager (accacceble via Esc key).
I wonder why it displayed splash screen at all when I was booting from HDD as it doesn't contain ESP partition with efi loader.
comment:8 by , 7 months ago
So, the serial log indicates we get all the way to the call to ExitBootServices. But then nothing else happens.
I added a bit more tracing. It appears that ExitBootServices is called at least twice; and after the second call, we don't get any more debug output. I tried attaching a debugger to VMware's GDB stub, but I can't really tell what's going on. It looks like execution may be stopped on a HLT instruction, but where or in what code I can't tell (based on the addresses it looks like before kernel handoff, though.)
A change to UEFI loader was revert in hrev56377. Was it better with a hrev56376?