0.24.2-80 x86 boots x86_64 doesn't

Help with booting MiniMyth

Moderator: Pablo

Post Reply
tsjoender
Contributor
Posts: 87
Joined: Mon Nov 05, 2007 9:45 am
Location: Friesland, The Netherlands

0.24.2-80 x86 boots x86_64 doesn't

Post by tsjoender » Mon Apr 23, 2012 5:06 pm

When I tried updating to 0.24.2-80 I could not get the x86_64 version to boot. It would throw a "Kernel panic - not syncing: No init found." message right after downloading the rootfs. I tried increasing the ramdisk size but that didn't help. The MD5 checksum of my download matches the one on the MiniMyth download page, so I have not tried downloading it again.

When I tried the 32bit variant of 0.24.2-80 it booted right away without issue.

For me it doesn't really matter if I use the 32bit or the 64 bit variant, so for now I'm sorted, but maybe the package available for download is corrupted or incomplete? Any others notice this?

Pablo
Site Admin
Posts: 4182
Joined: Tue Dec 14, 2004 2:13 am
Location: La Jolla
Contact:

Re: 0.24.2-80 x86 boots x86_64 doesn't

Post by Pablo » Tue Apr 24, 2012 3:59 am

Yes, there is something broken. Until I have time to track it down, I have removed the x86_64 images from the download directory.
MiniMyth running on an Acer ApireRevo 3610 and a Zotac ZBOX-ID80-U. Find out more at my MythTV page.

cheesy
Member
Posts: 16
Joined: Tue Apr 24, 2012 3:24 am

Re: 0.24.2-80 x86 boots x86_64 doesn't

Post by cheesy » Tue Apr 24, 2012 4:02 am

I ran into exactly the same problem and spent some time figuring it out.

It appears that there's a missing symlink in the rootfs; /lib64 -> rootfs-ro/lib64. I rebuilt the squashfs with that added and now my x86_64 build seems to be working OK.

So in answer to your question, the rootfs is actually broken. If you understand the above you can fix it, but posting my fix here would just be silly since it's a ~130MB file for a tiny fix (and even bigger since I don't have the right tools to make the squashfs). Hopefully Pablo sees this and can update the image (and build scripts).

Pablo
Site Admin
Posts: 4182
Joined: Tue Dec 14, 2004 2:13 am
Location: La Jolla
Contact:

Re: 0.24.2-80 x86 boots x86_64 doesn't

Post by Pablo » Tue Apr 24, 2012 5:15 am

cheesy wrote:I ran into exactly the same problem and spent some time figuring it out.

It appears that there's a missing symlink in the rootfs; /lib64 -> rootfs-ro/lib64. I rebuilt the squashfs with that added and now my x86_64 build seems to be working OK.

So in answer to your question, the rootfs is actually broken. If you understand the above you can fix it, but posting my fix here would just be silly since it's a ~130MB file for a tiny fix (and even bigger since I don't have the right tools to make the squashfs). Hopefully Pablo sees this and can update the image (and build scripts).
Thank you for tracking this down.

It appears a patch I make to gcc so that it will look for the dynamic linker in /lib rather than /lib64 broke with the latest upgrade of gcc. I have fixed it and I am rebuilding.
MiniMyth running on an Acer ApireRevo 3610 and a Zotac ZBOX-ID80-U. Find out more at my MythTV page.

cheesy
Member
Posts: 16
Joined: Tue Apr 24, 2012 3:24 am

Re: 0.24.2-80 x86 boots x86_64 doesn't

Post by cheesy » Thu Apr 26, 2012 10:28 pm

I notice that you have not yet posted updated builds. Do you need volunteers for testing?

Pablo
Site Admin
Posts: 4182
Joined: Tue Dec 14, 2004 2:13 am
Location: La Jolla
Contact:

Re: 0.24.2-80 x86 boots x86_64 doesn't

Post by Pablo » Thu Apr 26, 2012 11:27 pm

cheesy wrote:I notice that you have not yet posted updated builds. Do you need volunteers for testing?
I have a couple more things that I am fixing (e.g. saving theme downloaded using theme chooser and remote control key mappings). Once I upload it, I would appreciate having people test it.
MiniMyth running on an Acer ApireRevo 3610 and a Zotac ZBOX-ID80-U. Find out more at my MythTV page.

cheesy
Member
Posts: 16
Joined: Tue Apr 24, 2012 3:24 am

Re: 0.24.2-80 x86 boots x86_64 doesn't

Post by cheesy » Thu Apr 26, 2012 11:43 pm

Ahh, I guess my hacked-up 64-bit version isn't quite safe then ;) I'll keep an eye on this thread, thanks!

Pablo
Site Admin
Posts: 4182
Joined: Tue Dec 14, 2004 2:13 am
Location: La Jolla
Contact:

Re: 0.24.2-80 x86 boots x86_64 doesn't

Post by Pablo » Sun Apr 29, 2012 2:38 pm

I have uploaded a new test build. Could you test it?
MiniMyth running on an Acer ApireRevo 3610 and a Zotac ZBOX-ID80-U. Find out more at my MythTV page.

cheesy
Member
Posts: 16
Joined: Tue Apr 24, 2012 3:24 am

Re: 0.24.2-80 x86 boots x86_64 doesn't

Post by cheesy » Sun Apr 29, 2012 11:58 pm

Just dropped 0.24.3-81b3-x86_64 on my system. Booted just fine first time and no problems with my ir-remote. In fact the UI seems a little more responsive, but maybe that's just me.

Thanks!

dagnygren
Member
Posts: 20
Joined: Sun Feb 20, 2011 12:28 pm

Crash in 81b3

Post by dagnygren » Mon Apr 30, 2012 6:35 am

Got the following crash on my Zotac with 0.24.3-81.b3-x86_64:

Code: Select all

Apr 29 23:58:44 zotacsleep kernel: Modules linked in: cpufreq_userspace mperf uinput nvidia(PO) coretemp p4_clockmod freq_table spe
edstep_lib snd_hda_codec_hdmi hid_twinhan evdev snd_hda_codec_realtek i2c_i801 usbhid i2c_core hid snd_hda_intel psmouse snd_hda_co
Apr 29 23:58:44 zotacsleep kernel: Pid: 2025, comm: mythfrontend Tainted: P           O 3.3.4 #1 To Be Filled By O.E.M. To Be Fille
d By O.E.M./NM10
Apr 29 23:58:44 zotacsleep kernel: RIP: 0010:[<ffffffff8116f156>]  [<ffffffff8116f156>] inflate_fast+0xd2/0x540
Apr 29 23:58:44 zotacsleep kernel: RSP: 0018:ffff880074ec58a8  EFLAGS: 00010282
Apr 29 23:58:44 zotacsleep kernel: RAX: ffff880074ef6dac RBX: 0000000000000008 RCX: 0000000000000008
Apr 29 23:58:44 zotacsleep kernel: RDX: 0000000000000016 RSI: 0000000011713dc8 RDI: ffff880074ef6cec
Apr 29 23:58:44 zotacsleep kernel: RBP: ffffc90011713d00 R08: 000000000000002c R09: ffff88004f52d0c1
Apr 29 23:58:44 zotacsleep kernel: R10: ffffc90011713000 R11: 0000000000005ced R12: 0000000000000000
Apr 29 23:58:44 zotacsleep kernel: R13: ffffc90011715548 R14: 0000000000000400 R15: ffffc90011715547
Apr 29 23:58:44 zotacsleep kernel: FS:  00007ff59c18e700(0000) GS:ffff880077a80000(0000) knlGS:0000000000000000
Apr 29 23:58:44 zotacsleep kernel: CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
Apr 29 23:58:44 zotacsleep kernel: CR2: 00007ff58e4df000 CR3: 0000000078992000 CR4: 00000000000006e0
Apr 29 23:58:44 zotacsleep kernel: DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
Apr 29 23:58:44 zotacsleep kernel: DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
Apr 29 23:58:44 zotacsleep kernel: Process mythfrontend (pid: 2025, threadinfo ffff880074ec4000, task ffff880064241100)
Apr 29 23:58:44 zotacsleep kernel: Stack:
Apr 29 23:58:44 zotacsleep kernel:  000000000000001f 0000dced00008000 0000000000005ced ffffc90011713000
Apr 29 23:58:44 zotacsleep kernel:  0000000000000000 0000000000000000 ffff88004f52d000 0000000000000400
Apr 29 23:58:44 zotacsleep kernel:  ffff880074ecc9c0 ffffffff8116ff27 ffff880000000d8d ffffc90000000000
Apr 29 23:58:44 zotacsleep kernel: Call Trace:
Apr 29 23:58:44 zotacsleep kernel:  [<ffffffff8116ff27>] ? zlib_inflate+0x7d2/0xfb4
Apr 29 23:58:44 zotacsleep kernel:  [<ffffffff81117841>] ? zlib_uncompress+0x12d/0x1cb
Apr 29 23:58:44 zotacsleep kernel:  [<ffffffff81114a3c>] ? squashfs_read_data+0x350/0x450
Apr 29 23:58:44 zotacsleep kernel:  [<ffffffff81114cac>] ? squashfs_cache_get+0x170/0x290
Apr 29 23:58:44 zotacsleep kernel:  [<ffffffff81115167>] ? squashfs_read_metadata+0x97/0xc3
Apr 29 23:58:44 zotacsleep kernel:  [<ffffffff81115c73>] ? squashfs_readpage+0x4d5/0x746
Apr 29 23:58:44 zotacsleep kernel:  [<ffffffff8106465c>] ? __do_page_cache_readahead+0x18d/0x1cd
Apr 29 23:58:44 zotacsleep kernel:  [<ffffffff810648d5>] ? ra_submit+0x19/0x1d
Apr 29 23:58:44 zotacsleep kernel:  [<ffffffff8105e03e>] ? generic_file_aio_read+0x2a7/0x61d
Apr 29 23:58:44 zotacsleep kernel:  [<ffffffff8108411c>] ? do_sync_read+0xa5/0xdd
Apr 29 23:58:44 zotacsleep kernel:  [<ffffffff81390d6e>] ? rwsem_down_failed_common+0xe3/0x115
Apr 29 23:58:44 zotacsleep kernel:  [<ffffffff81079462>] ? page_add_new_anon_rmap+0x62/0x83
Apr 29 23:58:44 zotacsleep kernel:  [<ffffffff810847dd>] ? vfs_read+0x97/0x11e
Apr 29 23:58:44 zotacsleep kernel:  [<ffffffffa00012a8>] ? unionfs_read+0xb5/0x144 [unionfs]
Apr 29 23:58:44 zotacsleep kernel:  [<ffffffff810847dd>] ? vfs_read+0x97/0x11e
Apr 29 23:58:44 zotacsleep kernel:  [<ffffffff810848a2>] ? sys_read+0x3e/0x70
Apr 29 23:58:44 zotacsleep kernel:  [<ffffffff813917e2>] ? system_call_fastpath+0x16/0x1b
Apr 29 23:58:44 zotacsleep kernel: Code: 24 20 41 8d 48 ff 48 89 4c 24 50 41 8d 4b ff 44 01 db 89 5c 24 5c 48 ff c1 48 89 4c 24 60 83 fa 0e 77 22 41 0f b6 59 01 45 0f b6 <41> 02 88 d1 49 83 c1 02 48 d3 e3 8d 4a 08 83 c2 10 49 d3 e0 49 
Apr 29 23:58:44 zotacsleep kernel: RIP  [<ffffffff8116f156>] inflate_fast+0xd2/0x540
Apr 29 23:58:44 zotacsleep kernel:  RSP <ffff880074ec58a8>
Apr 29 23:58:44 zotacsleep kernel: ---[ end trace c49e04d9aa2f2de7 ]---

This happened when pressing the OK button trying to select "media livrary" (Not sure about the exact wording in English, running it in Swedish here.)

cheesy
Member
Posts: 16
Joined: Tue Apr 24, 2012 3:24 am

Re: 0.24.2-80 x86 boots x86_64 doesn't

Post by cheesy » Sat May 05, 2012 4:11 pm

I tried changing my language to Swedish but I did not see that crash myself. From the stack trace it appears to be in squashfs, which would imply a corrupted initrd.

Is the crash repeatable?

tsjoender
Contributor
Posts: 87
Joined: Mon Nov 05, 2007 9:45 am
Location: Friesland, The Netherlands

Re: 0.24.2-80 x86 boots x86_64 doesn't

Post by tsjoender » Sun May 06, 2012 6:21 pm

Just want to report back that the recently added x86_64 builds (0.24.3-81b3) boot fine on my setup. Thanks for the prompt attention!

cheesy
Member
Posts: 16
Joined: Tue Apr 24, 2012 3:24 am

Re: 0.24.2-80 x86 boots x86_64 doesn't

Post by cheesy » Fri Aug 10, 2012 9:44 pm

I notice that there are currently no x86_64 builds available to download, even for the new stable builds.

Pablo, have you stopped supporting x86_64?

Post Reply