Re: gpxe and memdisk could not boot winpe disk image
ShenQiyu <[email protected]>
| Newsgroups | gmane.network.etherboot.user |
|---|---|
| Message-ID | <[email protected]> |
Dear all, I have more test information for my issue: 1. Use gpxe chain with pxelinux + memdisk, the winpe disk image could also boot successfully. 2. When I break memdisk just before the disk image boot, the e820 memory map is: SUCCESS: gpxe + pxelinux + memdisk http protocol 00000000 00094000 1 00094000 0000c000 2 000dc000 00024000 2 00100000 9d3a1000 1 ... ramdisk at 0x966a9000, length 0x06dd1c00 (to 965D1C00) Old dos memory at 0x94000(map says 0x94000), loading at 0x93400 1588:0xffff 15E801: 0x3c00 0x956a old: int13=e1960aef int15 = f000f859 int1e = f000efc7 new: int13=9340000a int15 = 9340037f int1e = f000efc7 FAILURE: gpxe + memdisk http protocol 00000000 00090800 1 00090800 0000f800 2 000dc000 00024000 2 00100000 9d3a1000 1 ... ramdisk at 0x8f800000, length 0x06dd1c00 (to 965D1C00) Old dos memory at 0x90800(map says 0x90800), loading at 0x8fc00 1588:0xffff 15E801: 0x3c00 0x8E80 old: int13=e1960aef int15 = f000f859 int1e = f000efc7 new: int13=8fc0000a int15 = 8fc0037f int1e = f000efc7 It could found that: 1. The first and second e820 mapping ranges are different. 2. The disk image load address are different 3. The old dos memory address is different, and the new int13 address To compare the new int13 address, it is found that for the FAILURE case: the new int13 address is IN THE RANGE OF ramdisk! Anyone could please help this issue? Thanks! Paul From: [email protected] To: [email protected]; [email protected] Subject: gpxe and memdisk could not boot winpe disk image Date: Sat, 19 Sep 2009 22:03:09 +0800 Dear all After failure of booting winpe 2.0 disk image in gpxe + memdisk with slam protocal. I have tried other protocol, tftp and http, the winpe boot still stuck before show the desktop(the pe image itself is loaded). Because pxelinux + memdisk works well, so I think it maybe a problem with gpxe or gpxe+memdisk. Also I test with winpe 2005 disk image, it still failed, but with more information. First the desktop background could show partially without upper row(strip) Then BSOD happened with the following message: << a problem has.... framebuf if this ... The device driver got stuck in an infinite loop.... Please check you hardw... Technical information : *** Stop: 0x000000EA (0x84c29a60,0x84F01c48,0x84c263e8,0x00000001) framebuf >> I google it, someone says it is related to bad memory, others are graphics card. I guest it may related to the memory management compatibility issue of gpxe with memdisk. So I want some hints to debug the issue, check e820 memory map, or ... ? BTW, occassionally, the boot may succeed one or twice one day. But most of them failed. Anyone who want to get the winpe disk image for test, please contact me. Regards! Paul Date: Fri, 18 Sep 2009 20:55:21 +0000 Subject: Re: Re: [Etherboot-discuss] how to script for SLAM From: [email protected] To: [email protected]; [email protected] Paul, If you're able to use PXELINUX to load WinPE via MEMDISK but not from gPXE, you may want to try chainloading PXELINUX with your modified gPXE binary and then loading MEMDISK and WinPE from PXELINUX under gPXE, which at that point *should* support the SLAM protocol. When PXELINUX is chainloaded with gPXE, gPXE's protocol support is extended to PXELINUX. There's a DHCP option that you can set for PXELINUX that determines the "TFTP Prefix" you see when booting a client, which you can set to "x-slam://slamserver" to force use of the SLAM protocol. I don't know which option it is off the top of my head, though. Also, someone else may be able to add some more information, as there may be some sort of issue with the use of SLAM itself. Though that's just a guess on my part. Good luck, Paul. -Andrew Bobulsky _________________________________________________________________ 录禄谩碌虏禄氓碌路陆拢驴麓⒚铆碌录卯msn禄楼露炉鹿娄拢隆 http://ditu.live.com/?form=TL&swm=1 ------------------------------------------------------------------------------ Come build with us! The BlackBerry® Developer Conference in SF, CA is the only developer event you need to attend this year. Jumpstart your developing skills, take BlackBerry mobile applications to market and stay ahead of the curve. Join us from November 9-12, 2009. Register now! http://p.sf.net/sfu/devconf _______________________________________________ Etherboot-discuss mailing list [email protected] https://lists.sourceforge.net/lists/listinfo/etherboot-discuss