Re: Boot failure - svn up from this morning
Hi, I had the same problems, however, there is one more regression after these changes. It stably reproduces if you use EFI_STAGING_SIZE. I have custom FreeBSD distributive which has a sufficiently large mfsroot which is loaded through UEFI mode. To solve the problem, it was suggested to increase this variable and rebuild /sys/boot: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=209944 Also, it has to be increased periodically for some reason: https://svnweb.freebsd.org/base/stable/11/sys/boot/efi/loader/copy.c?r1=305779=305778=305779 I've try to ask why than threatens to increase this parameter at once large (or make it dependent on RAM): https://lists.freebsd.org/pipermail/freebsd-hackers/2016-November/050142.html but there are no answer options. At the moment, on r315141 boot via UEFI is fixed without EFI_STAGING_SIZE. With EFI_STAGING_SIZE=768 i got regression after last changes in panic with follow message: failed to allocate staging area:9 failed to allocate stating area Failed to start image provided by ZFS (5) http://pasteboard.co/IvbhU9Ffu.jpg I believe that there mfsroot may be greater than 64MB soon or later. Also there are no problems with loading big mfsroot images when MBR method is used. PS: I have ZFS-on-root system. With EFI_STAGING_SIZE=768 I lived with constant CURRENT svnup/rebuilds for about a year. So I will be glad if you pay attention to this. PPS: How to reproduce: 1) EFI_STAGING_SIZE=768 in /etc/make.conf 2) make -C /sys/boot clean 3) make -C /sys/boot 4) make -C /sys/boot install 5) reboot Thanks! ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: Boot failure - svn up from this morning
On 7 Mar 2017, at 16:50, Chris H wrote: On Tue, 07 Mar 2017 16:27:59 -0330 "Jonathan Anderson"wrote Hi, On 5 Mar 2017, at 20:31, Chris H wrote: OK copying the boot.efi from the install DVD will only hose the system (EFI). Before I attempt to do the same thing... what do you mean by "hose the system"? Is the correct recovery path to build a new USB image with "make release" post-r314828? That was *my* bad. I inadvertently copied the *wrong* file (I was in a hurry). So. That said; if you move loader.efi aside -- say; to _loader.efi. Then copy loader.efi from the install DVD, or any other system that's not too much older, and is good. You should be fine. :-) Make sure the permissions are correct, after the copy. Sorry for the misinformation. Thanks for the clarification! I am back up and running now. Cheers, Jon -- jonat...@freebsd.org ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
RE: Boot failure - svn up from this morning
> From: Alex Deiter [mailto:alex.dei...@gmail.com] > Hello Dexuan, > This issue reproduced at least for 4 different HW platform > How can I help you resolve this issue ? > > The same result for r314862: Hi guys, Sorry, I had to commit a new patch (r314891) just now to really fix the off-by-one bug. Thank Alex for reporting the issue in my previous patch (r314828). Hope this issue is finally fixed this time! :-) Thanks, -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
RE: Boot failure - svn up from this morning
> From: Alex Deiter [mailto:alex.dei...@gmail.com] > Sent: Wednesday, March 8, 2017 09:39 > To: Dexuan Cui <de...@microsoft.com>; freebsd-current curr...@freebsd.org> > Cc: Chris H <bsd-li...@bsdforge.com>; AN <a...@neu.net>; Ngie Cooper > (yaneurabeya) <yaneurab...@gmail.com>; Michael Tuexen > <tue...@freebsd.org>; Roberto Rodriguez Jr <rob.rodz@gmail.com>; Guido > Falsi <m...@madpilot.net>; Warner Losh <i...@bsdimp.com>; Ultima > <ultima1...@gmail.com>; Sepherosa Ziehau <se...@freebsd.org> > Subject: Re: Boot failure - svn up from this morning > > Hello Dexuan, > > This issue reproduced at least for 4 different HW platform: > > Supermicro 6037R-TXRF > Supermicro A1SRM-2758F > Supermicro X9SCM-F > Gigabyte GA-C1037UN-EU > > How can I help you resolve this issue ? > > Thank you! > > The same result for r314862: Weird... :-( I'm looking at the code again. Will report back soon. ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: Boot failure - svn up from this morning
Hello Dexuan, This issue reproduced at least for 4 different HW platform: Supermicro 6037R-TXRF Supermicro A1SRM-2758F Supermicro X9SCM-F Gigabyte GA-C1037UN-EU How can I help you resolve this issue ? Thank you! The same result for r314862: >> FreeBSD EFI boot block Loader path: /boot/loader.efi Initializing modules: ZFS UFS Probing 16 block devices..*.+.. done ZFS found the following pools: zroot UFS found no partitions Consoles: EFI console Staging area's size is reduced: 16384 -> 64! Command line arguments: loader.efi Image base: 0x6dc65000 EFI version: 2.31 EFI Firmware: American Megatrends (rev 5.08) FreeBSD/amd64 EFI loader, Revision 1.1 EFI boot environment Loading /boot/defaults/loader.conf /boot/kernel/kernel text=0x9210f8 elf64_loadimage: read failed can't load file '/boot/kernel/kernel': input/output error can't load file '/boot/kernel/kernel': input/output error Hit [Enter] to boot immediately, or any other key for command prompt. Booting [/boot/kernel/kernel] in 2 seconds... Type '?' for a list of commands, 'help' for more detailed help. OK OK memmap Type Physical Virtual #Pages Attr BootServicesCode 0008 UC WC WT WB ConventionalMemory 8000 0027 UC WC WT WB BootServicesData 0002f000 0011 UC WC WT WB BootServicesCode 0004 0060 UC WC WT WB ConventionalMemory 0010 0100 UC WC WT WB BootServicesData 0020 0040 UC WC WT WB ConventionalMemory 0024 0003fd80 UC WC WT WB LoaderData 3ffc 0040 UC WC WT WB ConventionalMemory 4000 00029c65 UC WC WT WB LoaderData 69c65000 4000 UC WC WT WB LoaderCode 6dc65000 0071 UC WC WT WB LoaderData 6dcd6000 2171 UC WC WT WB LoaderCode 6fe47000 001d UC WC WT WB MemoryMappedIO 6fe64000 0014 UC WC WT WB BootServicesData 6fe78000 ec32 UC WC WT WB ConventionalMemory 7eaaa000 037b UC WC WT WB BootServicesCode 7ee25000 0285 UC WC WT WB Reserved 7f0aa000 0075 UC WC WT WB ConventionalMemory 7f11f000 00ae UC WC WT WB ACPIMemoryNVS 7f1cd000 027b UC WC WT WB RuntimeServicesData 7f448000 01a7 UC WC WT WB RuntimeServicesCode 7f5ef000 005c UC WC WT WB BootServicesData 7f64b000 01b5 UC WC WT WB ConventionalMemory 0001 0078 UC WC WT WB MemoryMappedIO e000 4000 UC MemoryMappedIO fed01000 0003 UC MemoryMappedIO fed08000 0001 UC MemoryMappedIO fed0c000 0004 UC MemoryMappedIO fed1c000 0001 UC MemoryMappedIO fef0 1100 UC OK Thank you! Alex Deiter alex.dei...@gmail.com > On 7 Mar 2017, at 06:45, Dexuan Cui <de...@microsoft.com> wrote: > >> From: Dexuan Cui >> Sent: Tuesday, March 7, 2017 11:18 >> To: 'Alex Deiter' <alex.dei...@gmail.com> >> Cc: Chris H <bsd-li...@bsdforge.com>; AN <a...@neu.net>; Ngie Cooper >> (yaneurabeya) <yaneurab...@gmail.com>; Michael Tuexen >> <tue...@freebsd.org>; Roberto Rodriguez Jr <rob.rodz@gmail.com>; Guido >> Falsi <m...@madpilot.net>; Warner Losh <i...@bsdimp.com>; Ultima >> <ultima1...@gmail.com>; Sepherosa Ziehau <se...@freebsd.org> >> Subject: RE: Boot failure - svn up from this morning >> >>> From: Alex Deiter [mailto:alex.dei...@gmail.com] >>> Sent: Tuesday, March 7, 2017 11:04 >>> To: Dexuan Cui <de...@microsoft.com> >> >> >> Hi Alex, >> Thanks very much for the quick reply! >> >> I found an off-by-one bug in my patch and here is the fix. >> I'll commit it shortly. > > Hi Alex, > I committed Revision 314828 for this: > https://svnweb.freebsd.org/base?view=revision=314828 > > I believe it should fix the issue for you now. :-) > > Thanks, > -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: Boot failure - svn up from this morning
On Tue, 07 Mar 2017 16:27:59 -0330 "Jonathan Anderson"wrote > Hi, > > On 5 Mar 2017, at 20:31, Chris H wrote: > > > OK copying the boot.efi from the install DVD will only > > hose the system (EFI). > > Before I attempt to do the same thing... what do you mean by "hose the > system"? Is the correct recovery path to build a new USB image with > "make release" post-r314828? That was *my* bad. I inadvertently copied the *wrong* file (I was in a hurry). So. That said; if you move loader.efi aside -- say; to _loader.efi. Then copy loader.efi from the install DVD, or any other system that's not too much older, and is good. You should be fine. :-) Make sure the permissions are correct, after the copy. Sorry for the misinformation. --Chris > > Thanks, > > > Jon > -- > jonat...@freebsd.org ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: Boot failure - svn up from this morning
Hi, On 5 Mar 2017, at 20:31, Chris H wrote: OK copying the boot.efi from the install DVD will only hose the system (EFI). Before I attempt to do the same thing... what do you mean by "hose the system"? Is the correct recovery path to build a new USB image with "make release" post-r314828? Thanks, Jon -- jonat...@freebsd.org ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: Boot failure - svn up from this morning
On Mon, 6 Mar 2017 04:01:04 + Dexuan Cuiwrote > > From: owner-freebsd-curr...@freebsd.org [mailto:owner-freebsd- > > curr...@freebsd.org] On Behalf Of Dexuan Cui > > Hi Chris, > > Thank you very much for the screenshots!!! > > > > On the host there is a 1MB LoaderData memory range, which splits > > the big Conventional Memory range into a small one (15MB) and a > > big one: the small one is too small to hold the staging area. > > > > I'm going to post a patch shortly. > > Can you please try the below patch? > https://reviews.freebsd.org/D9904 > You can find the URL of the "Download Raw Diff" in the page and > 'wget' the patch and then apply it. > > It should be able to fix the recent UEFI-boot issue introduced by me. > > You may not need to re-buildworld. Please this link to replace the > bad 'loader.efi': > https://lists.freebsd.org/pipermail/freebsd-current/2017-March/064979.html > > I'm planning to commit the patch later today in about 6 hours, because > I'm pretty confident in the patch and it should fix the critical issue... It may already be too late. But I just wanted to let you know that I hadn't overlooked/discarded your message. $WORK$ is riding my a$$ pretty hard, and I'm not going to get a chance to catch my breath till at least mid day. But I *do* plan to give this a shot, and my earliest opportunity. Thanks, Dexuan! --Chris > > Thanks, > -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
RE: Boot failure - svn up from this morning
> From: Dexuan Cui > I committed r314770 just now to minimize the impact: > https://svnweb.freebsd.org/base?view=revision=314770 > > Please let me know in case this can't solve the issue. Sorry, r314770 has a bug, so I had to commit r314828 for this: https://svnweb.freebsd.org/base?view=revision=314828 Please make sure you use the latest code, i.e. >= r314828. Thanks, -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
RE: Boot failure - svn up from this morning
> From: owner-freebsd-curr...@freebsd.org [mailto:owner-freebsd- > curr...@freebsd.org] On Behalf Of Dexuan Cui > Hi Chris, > Thank you very much for the screenshots!!! > > On the host there is a 1MB LoaderData memory range, which splits > the big Conventional Memory range into a small one (15MB) and a > big one: the small one is too small to hold the staging area. > > I'm going to post a patch shortly. Can you please try the below patch? https://reviews.freebsd.org/D9904 You can find the URL of the "Download Raw Diff" in the page and 'wget' the patch and then apply it. It should be able to fix the recent UEFI-boot issue introduced by me. You may not need to re-buildworld. Please this link to replace the bad 'loader.efi': https://lists.freebsd.org/pipermail/freebsd-current/2017-March/064979.html I'm planning to commit the patch later today in about 6 hours, because I'm pretty confident in the patch and it should fix the critical issue... Thanks, -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
RE: Boot failure - svn up from this morning
> From: Dexuan Cui > Sent: Monday, March 6, 2017 13:10 > Can you please try the below patch? > https://reviews.freebsd.org/D9904 > You can find the URL of the "Download Raw Diff" in the page and > 'wget' the patch and then apply it. > > It should be able to fix the recent UEFI-boot issue introduced by me. > > You may not need to re-buildworld. Please this link to replace the > bad 'loader.efi': > https://lists.freebsd.org/pipermail/freebsd-current/2017-March/064979.html > > I'm planning to commit the patch later today in about 6 hours, because > I'm pretty confident in the patch and it should fix the critical issue... I committed r314770 just now to minimize the impact: https://svnweb.freebsd.org/base?view=revision=314770 Please let me know in case this can't solve the issue. Thanks, -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: Boot failure - svn up from this morning
On Mon, 6 Mar 2017 04:01:04 + Dexuan Cuiwrote > > From: owner-freebsd-curr...@freebsd.org [mailto:owner-freebsd- > > curr...@freebsd.org] On Behalf Of Dexuan Cui > > Hi Chris, > > Thank you very much for the screenshots!!! > > > > On the host there is a 1MB LoaderData memory range, which splits > > the big Conventional Memory range into a small one (15MB) and a > > big one: the small one is too small to hold the staging area. > > > > I'm going to post a patch shortly. > > Can you please try the below patch? Yes. I can't do it right away. Because that box is in the middle of a big build session. But the moment it's done. I'll give it a try, and report the results. Thanks again, Dexuan, for all your continued dedication to this! --Chris > https://reviews.freebsd.org/D9904 > You can find the URL of the "Download Raw Diff" in the page and > 'wget' the patch and then apply it. > > It should be able to fix the recent UEFI-boot issue introduced by me. > > You may not need to re-buildworld. Please this link to replace the > bad 'loader.efi': > https://lists.freebsd.org/pipermail/freebsd-current/2017-March/064979.html > > I'm planning to commit the patch later today in about 6 hours, because > I'm pretty confident in the patch and it should fix the critical issue... > > Thanks, > -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: Boot failure - svn up from this morning
On Mon, 6 Mar 2017 03:00:20 + Dexuan Cuiwrote > > From: Chris H [mailto:bsd-li...@bsdforge.com] > > Sent: Monday, March 6, 2017 09:57 > > > Thanks! I'm eager to see your screenshots. > > > The line whose "Physical" address contains 2MB is the most interesting to > > > me. And please at least post the other lines around the line. > > OK. Her you go. It's taken me some time to get any shots > > that are readable -- I don't have a very steady hand. :-( > > Anyway, I haven't touched them. They're just as my phone camera > > produced them. Because of their size, I've packed them all up. > > I'm afraid I don't know their exact order. Hopefully you'll > > know by looking at them. :-) > > They're located at: bsdforge.com/efi-memmap.tar.xz > > Hi Chris, > Thank you very much for the screenshots!!! No. Thank YOU, for all the work you're doing on this. Not to mention all the (personal) help you've given me. Greatly appreciated, Dexuan! --Chris > > On the host there is a 1MB LoaderData memory range, which splits > the big Conventional Memory range into a small one (15MB) and a > big one: the small one is too small to hold the staging area. > > I'm going to post a patch shortly. > > For people who are interested in the details: please see > https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=211746#c22 > > Thanks, > -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
RE: Boot failure - svn up from this morning
> From: owner-freebsd-curr...@freebsd.org [mailto:owner-freebsd- > curr...@freebsd.org] On Behalf Of Dexuan Cui > Hi Chris, > Thank you very much for the screenshots!!! > > On the host there is a 1MB LoaderData memory range, which splits > the big Conventional Memory range into a small one (15MB) and a > big one: the small one is too small to hold the staging area. > > I'm going to post a patch shortly. (Some of you may receive this same mail twice or even thrice -- sorry for the confusion, it's because my first mail "is being held until the list moderator can review it for approval" due to "Too many recipients to the message"...) Can you please try the below patch? https://reviews.freebsd.org/D9904 You can find the URL of the "Download Raw Diff" in the page and 'wget' the patch and then apply it. It should be able to fix the recent UEFI-boot issue introduced by me. You may not need to re-buildworld. Please this link to replace the bad 'loader.efi': https://lists.freebsd.org/pipermail/freebsd-current/2017-March/064979.html I'm planning to commit the patch later today in about 6 hours, because I'm pretty confident in the patch and it should fix the critical issue... Thanks, -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
RE: Boot failure - svn up from this morning
> From: Chris H [mailto:bsd-li...@bsdforge.com] > Sent: Monday, March 6, 2017 09:57 > > Thanks! I'm eager to see your screenshots. > > The line whose "Physical" address contains 2MB is the most interesting to > > me. > > And please at least post the other lines around the line. > OK. Her you go. It's taken me some time to get any shots > that are readable -- I don't have a very steady hand. :-( > Anyway, I haven't touched them. They're just as my phone camera > produced them. Because of their size, I've packed them all up. > I'm afraid I don't know their exact order. Hopefully you'll > know by looking at them. :-) > They're located at: bsdforge.com/efi-memmap.tar.xz Hi Chris, Thank you very much for the screenshots!!! On the host there is a 1MB LoaderData memory range, which splits the big Conventional Memory range into a small one (15MB) and a big one: the small one is too small to hold the staging area. I'm going to post a patch shortly. For people who are interested in the details: please see https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=211746#c22 Thanks, -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: Boot failure - svn up from this morning
Thank you verymuch for the reply, Dexuan! On Mon, 6 Mar 2017 00:34:19 + Dexuan Cuiwrote > > From: Chris H [mailto:bsd-li...@bsdforge.com] > > > Hi Alex, > > > Thanks for the info! > > > Unluckily it looks the delay() in my patch didn't work here somehow, so > > > the screen scrolled up so quickly that we're unable to see the output > > > clearly... :-( > > > > > > Luckily we can use this new method: > > > When you reach the "OK " prompt in your screenshot > > > please run "memmap" command > > > and share the output with me. Please press space to show the complete > > > info and share the screenshots. > > > > > > E.g. in my side, the memmap command shows: > > I'm experiencing the same problem. > > I've taken several screen shots, but not sure I got all of them. > > I guess the first, and last will maybe the most important. I'll > > clean them up, and post a link to them. > > Thanks! I'm eager to see your screenshots. > The line whose "Physical" address contains 2MB is the most interesting to me. > And please at least post the other lines around the line. OK. Her you go. It's taken me some time to get any shots that are readable -- I don't have a very steady hand. :-( Anyway, I haven't touched them. They're just as my phone camera produced them. Because of their size, I've packed them all up. I'm afraid I don't know their exact order. Hopefully you'll know by looking at them. :-) They're located at: bsdforge.com/efi-memmap.tar.xz Thank you very much for all your time, and attention to this, Dexuan! And sorry for my frustration. --Chris > > > In the meantime, is there any way to get the system to boot again? > > Is it possible to simply boot the install DVD, and replace the > > efi image on the system, with the one on the install DVD. Or > > can the offending commit simply be reverted. > > > > --Chris > Yes, you can replace the broken system's bad /boot/loader.efi with a > good version. You can also revert the offending commit ("loader.efi: reduce > the size of the staging area if necessary"). > > Thanks, > -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
RE: Boot failure - svn up from this morning
> From: Chris H [mailto:bsd-li...@bsdforge.com] > OK copying the boot.efi from the install DVD will only > hose the system (EFI). Do you mean copying the boot.efi from the install DVD doesn't work? If so we need to build a good boot.efi with the CURRENT code + reverting the offending commit. > So how long till (u)efi is again supported on FreeBSD? > Sorry for the frustration. But getting a working FreeBSD > on this system has become an unusually long, and expensive > process, this time around -- even for CURRENT. > > --Chris I think EFI boot has been supported by FreeBSD for years and it's only broken since last Thursday by my patch... Sorry. Let's make it right soon, once I understand why the patch breaks it by checking the memory maps on your system. Thanks, -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
RE: Boot failure - svn up from this morning
> From: Chris H [mailto:bsd-li...@bsdforge.com] > > Hi Alex, > > Thanks for the info! > > Unluckily it looks the delay() in my patch didn't work here somehow, so the > > screen scrolled up so quickly that we're unable to see the output clearly... > > :-( > > > > Luckily we can use this new method: > > When you reach the "OK " prompt in your screenshot > > please run "memmap" command > > and share the output with me. Please press space to show the complete info > > and share the screenshots. > > > > E.g. in my side, the memmap command shows: > I'm experiencing the same problem. > I've taken several screen shots, but not sure I got all of them. > I guess the first, and last will maybe the most important. I'll > clean them up, and post a link to them. Thanks! I'm eager to see your screenshots. The line whose "Physical" address contains 2MB is the most interesting to me. And please at least post the other lines around the line. > In the meantime, is there any way to get the system to boot again? > Is it possible to simply boot the install DVD, and replace the > efi image on the system, with the one on the install DVD. Or > can the offending commit simply be reverted. > > --Chris Yes, you can replace the broken system's bad /boot/loader.efi with a good version. You can also revert the offending commit ("loader.efi: reduce the size of the staging area if necessary"). Thanks, -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: Boot failure - svn up from this morning
On Sun, 05 Mar 2017 15:14:52 -0800 "Chris H"wrote > On Sun, 5 Mar 2017 10:48:32 + Dexuan Cui wrote > > > > From: Alex Deiter > > > Sent: Sunday, March 5, 2017 03:32 > > > > > > Hello, > > > > > > Screenshot: boot with patched loader: > > > Video: boot with patched loader: > > > > Hi Alex, > > Thanks for the info! > > Unluckily it looks the delay() in my patch didn't work here somehow, so the > > screen scrolled up so quickly that we're unable to see the output > > clearly... :-( > > > > Luckily we can use this new method: > > When you reach the "OK " prompt in your screenshot > > (http://picpaste.com/IMG_1768-PnmZAtBZ.JPG), please run "memmap" command > > and share the output with me. Please press space to show the complete info > > and share the screenshots. > > > > E.g. in my side, the memmap command shows: > I'm experiencing the same problem. > I've taken several screen shots, but not sure I got all of them. > I guess the first, and last will maybe the most important. I'll > clean them up, and post a link to them. > > In the meantime, is there any way to get the system to boot again? > Is it possible to simply boot the install DVD, and replace the > efi image on the system, with the one on the install DVD. Or > can the offending commit simply be reverted. OK copying the boot.efi from the install DVD will only hose the system (EFI). So how long till (u)efi is again supported on FreeBSD? Sorry for the frustration. But getting a working FreeBSD on this system has become an unusually long, and expensive process, this time around -- even for CURRENT. Thanks, for all your time, and consideration! --Chris > > Thanks! > > --Chris > > > > > Thanks, > > -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: Boot failure - svn up from this morning
On Sun, 5 Mar 2017 10:48:32 + Dexuan Cuiwrote > > From: Alex Deiter > > Sent: Sunday, March 5, 2017 03:32 > > > > Hello, > > > > Screenshot: boot with patched loader: > > Video: boot with patched loader: > > Hi Alex, > Thanks for the info! > Unluckily it looks the delay() in my patch didn't work here somehow, so the > screen scrolled up so quickly that we're unable to see the output clearly... > :-( > > Luckily we can use this new method: > When you reach the "OK " prompt in your screenshot > (http://picpaste.com/IMG_1768-PnmZAtBZ.JPG), please run "memmap" command > and share the output with me. Please press space to show the complete info > and share the screenshots. > > E.g. in my side, the memmap command shows: I'm experiencing the same problem. I've taken several screen shots, but not sure I got all of them. I guess the first, and last will maybe the most important. I'll clean them up, and post a link to them. In the meantime, is there any way to get the system to boot again? Is it possible to simply boot the install DVD, and replace the efi image on the system, with the one on the install DVD. Or can the offending commit simply be reverted. Thanks! --Chris > > Thanks, > -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
RE: Boot failure - svn up from this morning
> From: Alex Deiter > Sent: Sunday, March 5, 2017 03:32 > > Hello, > > Screenshot: boot with patched loader: > Video: boot with patched loader: Hi Alex, Thanks for the info! Unluckily it looks the delay() in my patch didn't work here somehow, so the screen scrolled up so quickly that we're unable to see the output clearly... :-( Luckily we can use this new method: When you reach the "OK " prompt in your screenshot (http://picpaste.com/IMG_1768-PnmZAtBZ.JPG), please run "memmap" command and share the output with me. Please press space to show the complete info and share the screenshots. E.g. in my side, the memmap command shows: Type Physical Virtual #Pages Attr ConventionalMemory 0080 UC WC WT WB BootServicesData 0008 0001 UC WC WT WB ConventionalMemory 00081000 001f UC WC WT WB BootServicesData 0010 0020 UC WC WT WB ConventionalMemory 0012 2e53 UC WC WT WB BootServicesData 02f73000 118d UC WC WT WB ConventionalMemory 0410 00037f00 UC WC WT WB LoaderData 3c00 4000 UC WC WT WB ConventionalMemory 4000 000b0adf UC WC WT WB LoaderData f0adf000 4000 UC WC WT WB LoaderCode f4adf000 0072 UC WC WT WB LoaderData f4b51000 217b UC WC WT WB LoaderCode f6ccc000 001d UC WC WT WB ACPIReclaimMemory f6ce9000 0001 UC WC WT WB Reserved f6cea000 0007 UC WC WT WB ACPIReclaimMemory f6cf1000 0001 UC WC WT WB RuntimeServicesData f6cf2000 0029 UC WC WT WB ConventionalMemory f6d1b000 0890 UC WC WT WB BootServicesData f75ab000 0017 UC WC WT WB ConventionalMemory f75c2000 0001 UC WC WT WB BootServicesData f75c3000 000d UC WC WT WB ConventionalMemory f75d 000c UC WC WT WB --more-- page down line down quit Thanks, -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: Boot failure - svn up from this morning
On Fri, Mar 03, 2017 at 02:30:08PM +0100, Michael Tuexen wrote: > > On 3 Mar 2017, at 12:59, Dexuan Cuiwrote: > > > >> From: Dexuan Cui > >> Sent: Friday, March 3, 2017 19:50 > >>> From: Alex Deiter > >>> Sent: Friday, March 3, 2017 17:22 > >>> Hello, > >>> The same issue with FreeBSD 12.0-CURRENT-r314563: > >>> elf64_loadimage: could not read symbols - skipped! > >>> > >>> I suspect regression after: > >>> > >>> Revision 314547 - Directory Listing > >>> Modified Thu Mar 2 07:25:50 2017 UTC (25 hours, 54 minutes ago) by dexuan > >>> loader.efi: reduce the size of the staging area if necessary > >> > >> Yes, I believe the issue is caused by the patch, which is supposed to PR > >> 211746: > >> ... > >> Can you please try the patch to dump the memory map? > >> ... > > > > BTW, I understand it's really annoying to boot the host first... > > I'm really sorry for this. > > > > I suppose you're able to build or find a good 'loader.efi' binary on > > another host, > > and then manage to replace the bad 'loader.efi' on the host broken by me. > > :-) > This problem also occurred on a Dell R430... And in bhyve with UEFI mode. -- Shawn Webb Cofounder and Security Engineer HardenedBSD GPG Key ID: 0x6A84658F52456EEE GPG Key Fingerprint: 2ABA B6BD EF6A F486 BE89 3D9E 6A84 658F 5245 6EEE signature.asc Description: PGP signature
Re: Boot failure - svn up from this morning
Hello, Screenshot: boot with patched loader: http://picpaste.com/IMG_1768-PnmZAtBZ.JPG Video: boot with patched loader: https://youtu.be/thzyRk0D36w Thank you! Alex Deiter alex.dei...@gmail.com > On 3 Mar 2017, at 16:37, Dexuan Cuiwrote: > >> From: Michael Tuexen >> Sent: Friday, March 3, 2017 21:30 >>> BTW, I understand it's really annoying to boot the host first... >>> I'm really sorry for this. >>> >>> I suppose you're able to build or find a good 'loader.efi' binary on >>> another host, >>> and then manage to replace the bad 'loader.efi' on the host broken by me. >>> :-) >> This problem also occurred on a Dell R430... >> >> Best regards >> Michael > > Can you please use the patch to dump the > : > https://lists.freebsd.org/pipermail/freebsd-current/2017-March/064979.html > > Let's get this fixed ASAP. > > Thanks, > -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: Boot failure - svn up from this morning
Hello Dexuan, Thank you for the patch! I’ll test it and let you know the result. Thank you! Alex Deiter alex.dei...@gmail.com > On 3 Mar 2017, at 14:49, Dexuan Cuiwrote: > >> From: Alex Deiter >> Sent: Friday, March 3, 2017 17:22 >> >> Hello, >> The same issue with FreeBSD 12.0-CURRENT-r314563: >> elf64_loadimage: could not read symbols - skipped! >> >> I suspect regression after: >> >> Revision 314547 - Directory Listing >> Modified Thu Mar 2 07:25:50 2017 UTC (25 hours, 54 minutes ago) by dexuan >> loader.efi: reduce the size of the staging area if necessary > > Yes, I believe the issue is caused by the patch, which is supposed to PR > 211746: > https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=211746 > > Sorry for causing the issue to you, but I suspect the patch reveals a bug in > your > host's firmware: the memory map reported by the host's firmware may be > incorrect. > > Can you please try the patch to dump the memory map? > https://github.com/dcui/freebsd/commit/6094aac8ac9bddb24e3ac45493ac020c94029ce8.patch > > And the patch will allow your host to boot. > Note: there is a 10-second pause every time 5 lines are printed. This is to > make > sure we have enough time to take a screenshot or photo. :-) > > About how to apply the patch and build/install it: > 'wget' the above patch, and 'cd' into your FREEBSD_SOURCE_ROOT/sys/boot/ and > run > "patch -p3 < the_patch_name". If you have run 'make buildworld", just run > 'make' in the > sys/boot/ directory and copy the new loader.efi into the boot folder, e.g. in > my side, I > use > cp -iv /usr/obj/root/bsd.git/sys/boot/efi/loader/loader.efi /boot/loader.efi > > Thanks, > -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: Boot failure - svn up from this morning
> On 3 Mar 2017, at 12:59, Dexuan Cuiwrote: > >> From: Dexuan Cui >> Sent: Friday, March 3, 2017 19:50 >>> From: Alex Deiter >>> Sent: Friday, March 3, 2017 17:22 >>> Hello, >>> The same issue with FreeBSD 12.0-CURRENT-r314563: >>> elf64_loadimage: could not read symbols - skipped! >>> >>> I suspect regression after: >>> >>> Revision 314547 - Directory Listing >>> Modified Thu Mar 2 07:25:50 2017 UTC (25 hours, 54 minutes ago) by dexuan >>> loader.efi: reduce the size of the staging area if necessary >> >> Yes, I believe the issue is caused by the patch, which is supposed to PR >> 211746: >> ... >> Can you please try the patch to dump the memory map? >> ... > > BTW, I understand it's really annoying to boot the host first... > I'm really sorry for this. > > I suppose you're able to build or find a good 'loader.efi' binary on another > host, > and then manage to replace the bad 'loader.efi' on the host broken by me. :-) This problem also occurred on a Dell R430... Best regards Michael > > Thanks, > -- Dexuan > ___ > freebsd-current@freebsd.org mailing list > https://lists.freebsd.org/mailman/listinfo/freebsd-current > To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org" smime.p7s Description: S/MIME cryptographic signature
RE: Boot failure - svn up from this morning
> From: Michael Tuexen > Sent: Friday, March 3, 2017 21:30 > > BTW, I understand it's really annoying to boot the host first... > > I'm really sorry for this. > > > > I suppose you're able to build or find a good 'loader.efi' binary on > > another host, > > and then manage to replace the bad 'loader.efi' on the host broken by me. > > :-) > This problem also occurred on a Dell R430... > > Best regards > Michael Can you please use the patch to dump the EFI memory map: https://lists.freebsd.org/pipermail/freebsd-current/2017-March/064979.html Let's get this fixed ASAP. Thanks, -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
RE: Boot failure - svn up from this morning
> From: Dexuan Cui > Sent: Friday, March 3, 2017 19:50 > > From: Alex Deiter > > Sent: Friday, March 3, 2017 17:22 > > Hello, > > The same issue with FreeBSD 12.0-CURRENT-r314563: > > elf64_loadimage: could not read symbols - skipped! > > > > I suspect regression after: > > > > Revision 314547 - Directory Listing > > Modified Thu Mar 2 07:25:50 2017 UTC (25 hours, 54 minutes ago) by dexuan > > loader.efi: reduce the size of the staging area if necessary > > Yes, I believe the issue is caused by the patch, which is supposed to PR > 211746: > ... > Can you please try the patch to dump the memory map? > ... BTW, I understand it's really annoying to boot the host first... I'm really sorry for this. I suppose you're able to build or find a good 'loader.efi' binary on another host, and then manage to replace the bad 'loader.efi' on the host broken by me. :-) Thanks, -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
RE: Boot failure - svn up from this morning
> From: Alex Deiter > Sent: Friday, March 3, 2017 17:22 > > Hello, > The same issue with FreeBSD 12.0-CURRENT-r314563: > elf64_loadimage: could not read symbols - skipped! > > I suspect regression after: > > Revision 314547 - Directory Listing > Modified Thu Mar 2 07:25:50 2017 UTC (25 hours, 54 minutes ago) by dexuan > loader.efi: reduce the size of the staging area if necessary Yes, I believe the issue is caused by the patch, which is supposed to PR 211746: https://bugs.freebsd.org/bugzilla/show_bug.cgi?id=211746 Sorry for causing the issue to you, but I suspect the patch reveals a bug in your host's firmware: the memory map reported by the host's firmware may be incorrect. Can you please try the patch to dump the memory map? https://github.com/dcui/freebsd/commit/6094aac8ac9bddb24e3ac45493ac020c94029ce8.patch And the patch will allow your host to boot. Note: there is a 10-second pause every time 5 lines are printed. This is to make sure we have enough time to take a screenshot or photo. :-) About how to apply the patch and build/install it: 'wget' the above patch, and 'cd' into your FREEBSD_SOURCE_ROOT/sys/boot/ and run "patch -p3 < the_patch_name". If you have run 'make buildworld", just run 'make' in the sys/boot/ directory and copy the new loader.efi into the boot folder, e.g. in my side, I use cp -iv /usr/obj/root/bsd.git/sys/boot/efi/loader/loader.efi /boot/loader.efi Thanks, -- Dexuan ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: Boot failure - svn up from this morning
Hello, The same issue with FreeBSD 12.0-CURRENT-r314563: elf64_loadimage: could not read symbols - skipped! http://picpaste.com/IMG_1764-vfZl1l5o.JPG I suspect regression after: Revision 314547 - Directory Listing Modified Thu Mar 2 07:25:50 2017 UTC (25 hours, 54 minutes ago) by dexuan loader.efi: reduce the size of the staging area if necessary The loader assumes physical memory in [2MB, 2MB + EFI_STAGING_SIZE) is Conventional Memory, but actually it may not, e.g. in the case of Hyper-V Generation-2 VM (i.e. UEFI VM) running on Windows Server 2012 R2 host, there is a BootServiceData memory block at the address 47.449MB and the memory is not writable. Without the patch, the loader will crash in efi_copy_finish(): see PR 211746. The patch verifies the end of the staging area, and reduces its size if necessary. This way, the loader will not try to write into the BootServiceData memory any longer. Thank Marcel Moolenaar for helping me on this issue! The patch also allocates the staging area in the first 1GB memory. See the comment in the patch for this. PR: 211746 Reviewed by:marcel, kib, sephe Approved by:sephe (mentor) MFC after: 2 weeks Sponsored by: Microsoft Differential Revision: https://reviews.freebsd.org/D9686 Alex Deiter alex.dei...@gmail.com > On 3 Mar 2017, at 09:20, Chris Hwrote: > > On Thu, 2 Mar 2017 21:32:39 -0800 "Ngie Cooper (yaneurabeya)" > wrote > >>> On Mar 2, 2017, at 21:16, Chris H wrote: >>> >>> On Thu, 2 Mar 2017 21:56:38 -0500 (EST) AN wrote >>> Hi: I'm having a major problem after updating a 12-current machine today. After buildworld/kernel/install cycle on reboot I'm getting the following failure: "/boot/kernel/kernel text=0xb7716f data=0x100548+0x398358 elf64_loadimage: read failed can't load file '/boot/kernel/kernel': input/output error can't load file '/boot/kernel/kernel': input/output error OK boot kernel.old elf64_loadimage: read failed can't load file '/boot/kernel.old/kernel': input/output error can't load file '/boot/kernel.old/kernel': input/output error" I have never experienced this failure before, and don't know how to proceed. Any help recovering from this would be greatly appreciated. Thanks in advance. >>> >>> I don't suppose you could post the output of >>> uname -a >>> >>> and maybe a copy of dmesg(8) could you? >> >> That would be good, but I don’t know if it’s possible (the OP is noting >> that the boot is broken when executing loader(8)..). -Ngie > > Indeed. But that doesn't stop the OP from booting from the > install media. ;-) :-) > > --Chris > > > ___ > freebsd-current@freebsd.org mailing list > https://lists.freebsd.org/mailman/listinfo/freebsd-current > To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org" ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: Boot failure - svn up from this morning
On Thu, 2 Mar 2017 21:32:39 -0800 "Ngie Cooper (yaneurabeya)"wrote > > On Mar 2, 2017, at 21:16, Chris H wrote: > > > > On Thu, 2 Mar 2017 21:56:38 -0500 (EST) AN wrote > > > >> Hi: > >> > >> I'm having a major problem after updating a 12-current machine today. > >> After buildworld/kernel/install cycle on reboot I'm getting the following > >> failure: > >> > >> "/boot/kernel/kernel text=0xb7716f data=0x100548+0x398358 > >> elf64_loadimage: read failed > >> can't load file '/boot/kernel/kernel': input/output error > >> can't load file '/boot/kernel/kernel': input/output error > >> > >> OK boot kernel.old > >> elf64_loadimage: read failed > >> can't load file '/boot/kernel.old/kernel': input/output error > >> can't load file '/boot/kernel.old/kernel': input/output error" > >> > >> I have never experienced this failure before, and don't know how to > >> proceed. Any help recovering from this would be greatly appreciated. > >> Thanks in advance. > > > > I don't suppose you could post the output of > > uname -a > > > > and maybe a copy of dmesg(8) could you? > > That would be good, but I don’t know if it’s possible (the OP is noting > that the boot is broken when executing loader(8)..). -Ngie Indeed. But that doesn't stop the OP from booting from the install media. ;-) :-) --Chris ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Re: Boot failure - svn up from this morning
> On Mar 2, 2017, at 21:16, Chris Hwrote: > > On Thu, 2 Mar 2017 21:56:38 -0500 (EST) AN wrote > >> Hi: >> >> I'm having a major problem after updating a 12-current machine today. >> After buildworld/kernel/install cycle on reboot I'm getting the following >> failure: >> >> "/boot/kernel/kernel text=0xb7716f data=0x100548+0x398358 >> elf64_loadimage: read failed >> can't load file '/boot/kernel/kernel': input/output error >> can't load file '/boot/kernel/kernel': input/output error >> >> OK boot kernel.old >> elf64_loadimage: read failed >> can't load file '/boot/kernel.old/kernel': input/output error >> can't load file '/boot/kernel.old/kernel': input/output error" >> >> I have never experienced this failure before, and don't know how to >> proceed. Any help recovering from this would be greatly appreciated. >> Thanks in advance. > > I don't suppose you could post the output of > uname -a > > and maybe a copy of dmesg(8) could you? That would be good, but I don’t know if it’s possible (the OP is noting that the boot is broken when executing loader(8)..). -Ngie signature.asc Description: Message signed with OpenPGP using GPGMail
Re: Boot failure - svn up from this morning
On Thu, 2 Mar 2017 21:56:38 -0500 (EST) ANwrote > Hi: > > I'm having a major problem after updating a 12-current machine today. > After buildworld/kernel/install cycle on reboot I'm getting the following > failure: > > "/boot/kernel/kernel text=0xb7716f data=0x100548+0x398358 > elf64_loadimage: read failed > can't load file '/boot/kernel/kernel': input/output error > can't load file '/boot/kernel/kernel': input/output error > > OK boot kernel.old > elf64_loadimage: read failed > can't load file '/boot/kernel.old/kernel': input/output error > can't load file '/boot/kernel.old/kernel': input/output error" > > I have never experienced this failure before, and don't know how to > proceed. Any help recovering from this would be greatly appreciated. > Thanks in advance. I don't suppose you could post the output of uname -a and maybe a copy of dmesg(8) could you? --Chris ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"
Boot failure - svn up from this morning
Hi: I'm having a major problem after updating a 12-current machine today. After buildworld/kernel/install cycle on reboot I'm getting the following failure: "/boot/kernel/kernel text=0xb7716f data=0x100548+0x398358 elf64_loadimage: read failed can't load file '/boot/kernel/kernel': input/output error can't load file '/boot/kernel/kernel': input/output error OK boot kernel.old elf64_loadimage: read failed can't load file '/boot/kernel.old/kernel': input/output error can't load file '/boot/kernel.old/kernel': input/output error" I have never experienced this failure before, and don't know how to proceed. Any help recovering from this would be greatly appreciated. Thanks in advance. ___ freebsd-current@freebsd.org mailing list https://lists.freebsd.org/mailman/listinfo/freebsd-current To unsubscribe, send any mail to "freebsd-current-unsubscr...@freebsd.org"