[Letux-kernel] 5.4-rc1 woes

Andreas Kemnade andreas at kemnade.info
Wed Oct 2 22:19:31 CEST 2019


On Wed, 2 Oct 2019 21:18:12 +0200
Andreas Kemnade <andreas at kemnade.info> wrote:

> On Wed, 2 Oct 2019 21:01:29 +0200
> "H. Nikolaus Schaller" <hns at goldelico.com> wrote:
> 
> > > Am 02.10.2019 um 20:53 schrieb Andreas Kemnade <andreas at kemnade.info>:
> > > 
> > > Hi,
> > > 
> > > On Wed, 2 Oct 2019 20:44:27 +0200
> > > "H. Nikolaus Schaller" <hns at goldelico.com> wrote:
> > >     
> > >>> Am 02.10.2019 um 17:18 schrieb Andreas Kemnade <andreas at kemnade.info>:
> > >>> 
> > >>> The latter one first appeared in v5.3-10547-g45824fc0da6e      
> > >> 
> > >> hm -> mainly powerpc
> > >>     
> > >>> last test executed before was with
> > >>> v5.3-8029-g3c2edc36a774      
> > >>     
> > >> -> mainly pinmux      
> > >> 
> > >> but none of them obviously influences arm/omap.
> > >> 
> > >> Maybe they "fixed" something which reveals an old bug?
> > >> Like the bad pinctrl reference for the openpandora
> > >> (pinctrl was defined for a node that was part of the
> > >> other pinmux controller).    
> > > 
> > > ok, research state:
> > > 
> > > 1. panel-simple problem: fixed and pushed    
> > 
> > thanks! It was a warning and not upstream, right?
> >   
> well, no display. Well, whether it is upstream or not is a question of
> documentation+definition. I just fixed it on our side for simplicity
> in the dtb. Maybe other users of panel-simple are also affected.
> 
> > > 2. sporadic mmc stuff: unknown.    
> > 
> > this is the biggest problem...
> > AFAIR it was on GTA04 only and not on Pyra?
> >   
> no idea if it is the biggest problem. Not tested yet on pyra.
> It was mmc sdio:
> [   11.057006] refcount_t: underflow; use-after-free.
> [   11.062011] Modules linked in: bmp280_i2c(+) itg3200(+) bmp280 at24 hmc5843_i2c(+) hmc5843_core bma180(+) tsc2007(+) leds_tca6507 snd_soc_omap_mcbsp industrialio_triggered_buffer kfifo_buf snd_soc_ti_sdma phy_twl4030_usb(+) musb_hdrc snd_soc_twl4030 twl4030_vibra twl4030_madc gnss_sirf twl4030_charger twl4030_pwrbutton industrialio gnss ehci_omap(+) omapdss omapdss_base drm drm_panel_orientation_quirks cec overlay
> [   11.116027] usb usb1: Manufacturer: Linux 5.3.0-12397-gf41def397161 ehci_hcd
> [   11.123443] usb usb1: SerialNumber: 48064800.ehci
> [   11.163208] CPU: 0 PID: 76 Comm: kworker/0:2 Not tainted 5.3.0-12397-gf41def397161 #1
> [   11.171417] Hardware name: Generic OMAP36xx (Flattened Device Tree)
> [   11.177978] Workqueue: events_freezable mmc_rescan
> [   11.183013] [<c011028c>] (unwind_backtrace) from [<c010bd30>] (show_stack+0x10/0x14)
> [   11.191131] [<c010bd30>] (show_stack) from [<c07be0c0>] (dump_stack+0x7c/0x9c)
> [   11.198699] [<c07be0c0>] (dump_stack) from [<c012beb0>] (__warn+0xd4/0xf0)
> [   11.205902] [<c012beb0>] (__warn) from [<c012bf10>] (warn_slowpath_fmt+0x44/0x6c)
> [   11.213745] [<c012bf10>] (warn_slowpath_fmt) from [<c044c260>] (refcount_sub_and_test_checked+0x7c/0xa0)
> [   11.223693] [<c044c260>] (refcount_sub_and_test_checked) from [<c07c2528>] (kobject_put+0x38/0xbc)
> [   11.233062] [<c07c2528>] (kobject_put) from [<c066aff4>] (mmc_sdio_init_card+0x324/0xb28)
> [   11.241638] [<c066aff4>] (mmc_sdio_init_card) from [<c066ba7c>] (mmc_attach_sdio+0x84/0x320)
> [   11.250488] [<c066ba7c>] (mmc_attach_sdio) from [<c06635bc>] (mmc_rescan+0x318/0x33c)
> [   11.258666] [<c06635bc>] (mmc_rescan) from [<c0145008>] (process_one_work+0x280/0x460)
> [   11.266967] [<c0145008>] (process_one_work) from [<c0145d8c>] (worker_thread+0x28c/0x3ec)
> [   11.275512] [<c0145d8c>] (worker_thread) from [<c014a1d0>] (kthread+0x130/0x144)
> [   11.283264] [<c014a1d0>] (kthread) from [<c01010e8>] (ret_from_fork+0x14/0x2c)
> [   11.290802] Exception stack(0xdcf9dfb0 to 0xdcf9dff8)
> [   11.296081] dfa0:                                     00000000 00000000 00000000 00000000
> [   11.304626] dfc0: 00000000 00000000 00000000 00000000 00000000 00000000 00000000 00000000
> [   11.313201] dfe0: 00000000 00000000 00000000 00000000 00000013 00000000
> 
> 
> > > 3. bad adress errno: currently running git bisect run ...
> > > started with bad = 5.4-rc1
> > > good =  f41def397161053eb0d3ed6861ef65985efbf293
> > > 
> > > lets see if it reveals something interesting and if not,
> > > no wasted time since it runs automatic.    
> > 
> > Let's hope it finds something.
> >   
> at least it is the biggest problem to start debug anything else. At the
> moment I think it is a memory management one.
> 
andi at aktux:~/gta04/gta04-autotest-kernel$ git bisect log
git bisect start
# bad: [54ecb8f7028c5eb3d740bb82b0f1d90f2df63c5c] Linux 5.4-rc1
git bisect bad 54ecb8f7028c5eb3d740bb82b0f1d90f2df63c5c
# good: [f41def397161053eb0d3ed6861ef65985efbf293] Merge tag 'ceph-for-5.4-rc1' of git://github.com/ceph/ceph-client
git bisect good f41def397161053eb0d3ed6861ef65985efbf293
# bad: [e37e3bc7e265d05d00f14079767537699cf6bd46] Merge tag 'pwm/for-5.4-rc1' of git://git.kernel.org/pub/scm/linux/kernel/git/thierry.reding/linux-pwm
git bisect bad e37e3bc7e265d05d00f14079767537699cf6bd46
# bad: [a7b7b772bb4abaa4b2d9df67b50bf7208203da82] Merge branch 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip
git bisect bad a7b7b772bb4abaa4b2d9df67b50bf7208203da82
# bad: [972a2bf7dfe39ebf49dd47f68d27c416392e53b1] Merge tag 'nfs-for-5.4-1' of git://git.linux-nfs.org/projects/anna/linux-nfs
git bisect bad 972a2bf7dfe39ebf49dd47f68d27c416392e53b1
# bad: [b4ed71f557e458257e0f71b11969954acb389240] mm: treewide: clarify pgtable_page_{ctor,dtor}() naming
git bisect bad b4ed71f557e458257e0f71b11969954acb389240
# bad: [ac7c3e4ff401b304489a031938dbeaab585bfe0a] compiler: enable CONFIG_OPTIMIZE_INLINING forcibly
git bisect bad ac7c3e4ff401b304489a031938dbeaab585bfe0a
# good: [94fb98450456da82a16a378816390d99b85edb55] checkpatch: allow consecutive close braces
git bisect good 94fb98450456da82a16a378816390d99b85edb55
# good: [4fadcd1c14d810ec6a695039cfc71e03ae742deb] fs/reiserfs/fix_node.c: remove set but not used variables
git bisect good 4fadcd1c14d810ec6a695039cfc71e03ae742deb
# good: [8495f7e6732ed248b648d36439795b42ec650b9e] fork: improve error message for corrupted page tables
git bisect good 8495f7e6732ed248b648d36439795b42ec650b9e
# good: [7c3a6aedcd6aae0a32a527e68669f7dd667492d1] kexec: bail out upon SIGKILL when allocating memory.
git bisect good 7c3a6aedcd6aae0a32a527e68669f7dd667492d1
# good: [9dd819a15162f8f82a6001b090caa38c18297b39] uaccess: add missing __must_check attributes
git bisect good 9dd819a15162f8f82a6001b090caa38c18297b39
# first bad commit: [ac7c3e4ff401b304489a031938dbeaab585bfe0a] compiler: enable CONFIG_OPTIMIZE_INLINING forcibly

hmmmm...
Now it depends on compliers?!
andi at aktux:~/gta04/gta04-autotest-kernel$ arm-linux-gnueabihf-gcc --version
arm-linux-gnueabihf-gcc (Debian 8.3.0-2) 8.3.0
Copyright (C) 2018 Free Software Foundation, Inc.
This is free software; see the source for copying conditions.  There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.

And how to debug it? Well, lets first check whether reverting that
improves things on 5.4-rc1.

Regards,
Andreas


More information about the Letux-kernel mailing list