Hi Jurgen, > IIRC: Wasn't it impossible to boot into a standard > single user > with 6401605, without getting the corrupted text > display? > > A boot to milestone=none worked, though; but the text > display > would be corrupted as soon as "devfsadm" was > invoked? >
I remember that thread with you and Jason Fiset and at the time I tried and both answer were Yes. Just for testing I re-tried them today on an HP dv5000, on which I had to use the work around to install (I reboot from the CD five minutes ago to be sure). I also deleted the reference to pcie_error_disable_flag in my /etc/system and... surprise: pci bus 0x0000 cardnum 0x05 function 0x00: vendor 0x1002 device 0x5a37 ATI Technologies Inc Device unknown STATUS 0x0010 COMMAND 0x0040 CLASS 0x06 0x04 0x00 REVISION 0x00 HEADER 0x01 LATENCY 0x00 PRIBUS 0x00 SECBUS 0x02 SUBBUS 0x03 SECLT 0x00 SECSTATUS 0x0000 IOBASE 0x00000000 IOLIM 0x00000fff PREFETCH_MEMBASE 0xfff00000 MEMLIM 0x000fffff NO_FAST_B2B NO_SEC_BUS_RST NO_M_ABRT NO_VGA_EN NO_ISA_EN NO_SERR_EN PERR_EN I/O and memory access are disabled. So, unless I am really wrong, it looks like the bug has been silently fixed. I will look at the code tomorrow. Regards, J-F This message posted from opensolaris.org _______________________________________________ opensolaris-discuss mailing list opensolaris-discuss@opensolaris.org