Date: Sun, 6 Dec 2020 17:27:27 -0800 From: Marcel Flores <marcel@brickporch.com> To: mmel@freebsd.org Cc: freebsd-arm@freebsd.org Subject: Re: ThunderX Panic after r368370 Message-ID: <BFD08478-5312-4866-8EF5-6C4BCD8EED5E@brickporch.com> In-Reply-To: <7DFA7D8E-45A6-48B8-BB74-CC2EE29AF73C@brickporch.com> References: <1C3442ED-278E-45B8-9206-0DD24FCBC237@brickporch.com> <4331eee0-74a6-565c-3bec-0051415b2bc1@freebsd.org> <56F0E9EB-0B78-4B0B-830A-48F8AFC5ABE1@yahoo.com> <91654fc4-8734-d8a7-5309-0400f418438a@freebsd.org> <7DFA7D8E-45A6-48B8-BB74-CC2EE29AF73C@brickporch.com>
next in thread | previous in thread | raw e-mail | index | archive | help
> On Dec 6, 2020, at 4:59 PM, Marcel Flores <marcel@brickporch.com> = wrote: >> On Dec 6, 2020, at 3:51 AM, Michal Meloun <meloun.michal@gmail.com> = wrote: >>=20 >>=20 >>=20 >> On 06.12.2020 10:47, Mark Millard wrote: >>> On 2020-Dec-6, at 00:17, Michal Meloun <meloun.michal at gmail.com> = wrote: >>>> On 06.12.2020 3:21, Marcel Flores wrote: >>>>> Hi All, >>>>> Looks like the ThunderX started panicking at boot after r368370: >>>>> https://reviews.freebsd.org/rS368370 >>>>> =46rom a verbose boot, it looks like it bails in gic0 = redistributor setup(?): >>>>> gic0: CPU29 Re-Distributor woke up >>>>> gic0: CPU24 enabled CPU interface via system registers >>>>> gic0: CPU17 enabled CPU interface via system registers >>>>> gic0: CPU29 enabled CPU interface via system registers >>>>> done >>>>> Full Verbose boot: >>>>> https://gist.github.com/mesflores/f026122495c8494d041bce04d30b15bb >>>>> I'm not really familiar with the details of the commit, but happy = to test >>>>> anything if anyone has any ideas. >>>>=20 >>>>=20 >>>> Hi Marcel >>>> are you able to get crashdump and do backtrace? >>>> = https://www.freebsd.org/doc/en/books/developers-handbook/kerneldebug.html#= kerneldebug-obtain >>>> and >>>> = https://www.freebsd.org/doc/en/books/developers-handbook/kerneldebug-gdb.h= tml >>>> If not, I'll make some debug patch. >>>>=20 >>>> It's weird, even though GIC is potentially affected by my patch, in = this case the cpuid numbering was not changed. >>> (I've no access to a ThunderX. I just looked for my own curiosity. >>> Sorry if this is obvious and so is noise.) >>> When I looked at the code it appeared to be the last "->" in >>> the following that was dereferencing the nullptr value (via [x8] >>> in assembler notation): >>> static uint64_t >>> its_cmd_prepare(struct its_cmd *cmd, struct its_cmd_desc *desc) >>> { >>> uint64_t target; >>> uint8_t cmd_type; >>> u_int size; >>> cmd_type =3D desc->cmd_type; >>> target =3D ITS_TARGET_NONE; >>> switch (cmd_type) { >>> case ITS_CMD_MOVI: /* Move interrupt ID to another = collection */ >>> target =3D desc->cmd_desc_movi.col->col_target; >>> . . . >>> In other words: it appeared to me that the above = desc->cmd_desc_movi.col >>> evaluated as 0 when used in what was reported. >> This is very probably right analysis. But problem is that = cmd_desc_movi.col should not be NULL, is initialized in its_cmd_movi = from sc->sc_its_cols which should be allocated in gicv3_its_attach(). >>=20 >>=20 >> Marcel, can you, please also try this debug patch? >> = https://github.com/strejda/freebsd/commit/a25ed736644b05672e3e813891af213c= 280daac3 >> Unfortunately, I have only single socket board with GIv3, Honeycomb, = but it still boots fine. >>=20 >> Thanks, Michal >=20 > Debug patch output here (I also switched from GENERIC-NODEBUG to = GENERIC): >=20 > https://gist.github.com/mesflores/27bd1cca45b04e5b938166c9f1f79a04 >=20 > Having a little trouble getting the crashdump saved, but will update = if I can sort it out. >=20 > -m Here=E2=80=99s the debug build output when applied to r368369, incase = the side-by-side is helpful: https://gist.github.com/mesflores/329a826508dc47a0f9333ab5b6c52116 -m
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?BFD08478-5312-4866-8EF5-6C4BCD8EED5E>