Date: Wed, 2 May 2018 09:51:03 +0100 From: Mark Raynsford <list+org.freebsd.virtualization@io7m.com> To: Peter Grehan <grehan@freebsd.org> Cc: Fabian Freyer <fabian.freyer@physik.tu-berlin.de>, freebsd-virtualization@freebsd.org Subject: Re: Segmentation fault in grub-bhyve when trying to boot a Linux guest Message-ID: <20180502095103.3bec8a24@almond.int.arc7.info> In-Reply-To: <f67711f7-88ed-7757-dc26-98a4834486af@freebsd.org> References: <20180430230524.6c957e6c@almond.int.arc7.info> <AE2FBCC1-6E22-427B-96EB-5CE9CD898051@physik.tu-berlin.de> <20180501174142.31d7fc3a@almond.int.arc7.info> <f67711f7-88ed-7757-dc26-98a4834486af@freebsd.org>
next in thread | previous in thread | raw e-mail | index | archive | help
--Sig_/PV0CpmXYvA45+I4eKjscRlS Content-Type: text/plain; charset=US-ASCII Content-Transfer-Encoding: quoted-printable On 2018-05-01T18:06:38 -0700 Peter Grehan <grehan@freebsd.org> wrote: > > * rosemary_disk0.lzma (the LZMA compressed zvol) =20 >=20 > I was able to boot this image on a 12-current Ryzen system. Debian 9.4= =20 > also installed fine with the netinstall ISO and could boot. Bizarrely, I am also able to boot it without issue now. I worked through this with someone in the #bhyve channel on Freenode. Essentially: I couldn't boot the image with the stripped grub-bhyve binary, this would crash reliably. Restarting the hardware didn't make any difference, and destroying the vm after each attempt didn't make any difference either. I couldn't boot the image with the debug grub-bhyve binary. The same applied to that as above. If I ran the debug binary in gdb 8, the program ran to completion without crashing. After running it a few times without crashing in gdb, both the stripped and debug grub-bhyve binaries would run to completion *outside of the debugger* without crashing! Both the stripped and debug grub-bhyve binaries reliably work now. I can boot the VM, reboot it, whatever. I've restarted the hardware many times and cannot reproduce the original crash. I've never seen anything like this and have no theories as to why it didn't work at all, and now works reliably despite nothing apparently having changed anywhere. --=20 Mark Raynsford | http://www.io7m.com --Sig_/PV0CpmXYvA45+I4eKjscRlS Content-Type: application/pgp-signature Content-Description: OpenPGP digital signature -----BEGIN PGP SIGNATURE----- iHUEARYKAB0WIQS4Thd0dhbGF0xo1eVcGntxKBLMBQUCWul79wAKCRBcGntxKBLM BdhgAP9ekkk+Pz5j6Tm8eMnpCm0FKdLlQFMvkmuz+CixiQPPmAD+KHkGiWlGYo7N 3bQhn0VTT9JW93UXpBcwOZ7vTeaemAo= =tU5Z -----END PGP SIGNATURE----- --Sig_/PV0CpmXYvA45+I4eKjscRlS--
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?20180502095103.3bec8a24>