mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Greg Ungerer <gerg@linux-m68k.org>
To: Michael Schmitz <schmitzmic@gmail.com>,
	Geert Uytterhoeven <geert@linux-m68k.org>,
	Finn Thain <fthain@linux-m68k.org>
Cc: linux-m68k@lists.linux-m68k.org, linux-kernel@vger.kernel.org
Subject: Re: [PATCH RESEND] m68k: Handle put_user faults more carefully
Date: Thu, 8 Oct 2026 23:41:05 +1000	[thread overview]
Message-ID: <92bfc41e-3c37-4485-b633-f41f58f0b0bd@linux-m68k.org> (raw)
In-Reply-To: <d52fd72d-fe72-4fd3-9ae8-5ed999d18b5f@gmail.com>

Hi Michael,

On 8/10/26 09:54, Michael Schmitz wrote:
> Hi Greg,
> 
> thanks for running these tests!
> 
> On 6/10/26 01:42, Greg Ungerer wrote:
>> Hi Michael,
>>
>> On 3/10/26 09:55, Michael Schmitz wrote:
>>> Hi Geert,
>>>
>>> On 03/10/2026 5:28 AM, Geert Uytterhoeven wrote:
>>>
>>>> Hi Michael, Finn,
>>>>
>>>> On Thu, 1 Oct 2026 at 20:57, Michael Schmitz <schmitzmic@gmail.com> 
>>>> wrote:
>>>>> On 1/10/26 22:08, Finn Thain wrote:
>>>>>> Running 'stress-ng --sysbadaddr -1' on my MC68040 system immediately
>>>>>> produces an oops:
>>>>>>
>>>>>>        Unable to handle kernel access at virtual address f1c422fb
>>>>>>        Oops: 00000000
>>>>>>        Modules linked in:
>>>>>>        PC: [<00005a1a>] do_040writeback1+0xae/0x160
>>>>>>        SR: 2010  SP: 96087dc2  a2: 01500000
>>>>>>        d0: 00000000    d1: 014e8000    d2: 00000081    d3: 00000000
>>>>>>        d4: 00000481    d5: c043dfff    a0: 014e8000    a1: 00632fd5
>>>>>>        Process stress-ng (pid: 221, task=1b729d30)
>>>>>>        Frame format=7 eff addr=014e9eb8 ssw=0c81 faddr=c043dfff
>>>>>>        wb 1 stat/addr/data: 0001 00000481 c043dfff
>>>>>>        wb 2 stat/addr/data: 0081 c043dfff 2f746d70
>>>>>>        wb 3 stat/addr/data: 0005 014e9ed4 2f746d70
>>>>>>        push data: c043dfff 800daa70 00000000 014e9f1c
>>>>>>        Stack from 014e9ed4:
>>>>>>          2f740000 8017c000 014e9f10 00006830 00000081 c043dfff 
>>>>>> 2f746d70 2f746d70
>>>>>>          0081a000 00000400 c043dfff 800daa70 c043dfff 80016a20 
>>>>>> 00000003 014e9f88
>>>>>>          000026c8 014e9f1c 00000001 2f746d70 0081a000 00000400 
>>>>>> c043dfff 0081afff
>>>>>>          c043dfff 01500000 00000001 ffffffff 00000000 20000046 
>>>>>> 41d67008 014e9f58
>>>>>>          04810005 00810045 c043dfff 014e9f84 2f746d70 c043dfff 
>>>>>> 2f746d70 c043dfff
>>>>>>          80016a20 8017c000 00000000 0081affb 00000005 014e9fc4 
>>>>>> 00128bc6 c043dfff
>>>>>>        Call Trace: [<00006830>] buserr_c+0x510/0x698
>>>>>>          [<000026c8>] buserr+0x20/0x28
>>>>>>          [<00128bc6>] sys_getcwd+0xc8/0x15a
>>>>>>          [<000027a2>] syscall+0x8/0xc
>>>>>>          [<0008800d>] sanity_check_segment_list+0x17/0x11e
>>>>>>
>>>>>>        Code: 000c 0e90 1800 220f 0281 ffff e000 2041 <2228> 0008 
>>>>>> 0281 00ff
>>>>>>              ff00 67a4 6026 4280 122e 0013 206e 000c 0e10 1800 
>>>>>> 220f 0281
>>>>>>
>>>>>> 0000596c <do_040writeback1>:
>>>>>>        596c:       4e56 fffc       linkw %fp,#-4
>>>>>>        5970:       2f02            movel %d2,%sp@-
>>>>>>        5972:       202e 0008       movel %fp@(8),%d0
>>>>>>        5976:       4282            clrl %d2
>>>>>>        5978:       3400            movew %d0,%d2
>>>>>>        597a:       220f            movel %sp,%d1
>>>>>>        597c:       0281 ffff e000  andil #-8192,%d1
>>>>>>        5982:       2041            moveal %d1,%a0
>>>>>>        5984:       2228 0008       movel %a0@(8),%d1
>>>>>>        5988:       0281 00ff ff00  andil #16776960,%d1
>>>>>>        598e:       6600 0104       bnew 5a94 <do_040writeback1+0x128>
>>>>>>        5992:       4e7b 2000       movec %d2,%sfc
>>>>>>        5996:       4e7b 2001       movec %d2,%dfc
>>>>>>        599a:       0240 0060       andiw #96,%d0
>>>>>>        599e:       0c40 0020       cmpiw #32,%d0
>>>>>>        59a2:       6700 0084       beqw 5a28 <do_040writeback1+0xbc>
>>>>>>        59a6:       0c40 0040       cmpiw #64,%d0
>>>>>>        59aa:       6730            beqs 59dc <do_040writeback1+0x70>
>>>>>>        59ac:       4a40            tstw %d0
>>>>>>        59ae:       6752            beqs 5a02 <do_040writeback1+0x96>
>>>>>>        59b0:       4280            clrl %d0
>>>>>>        59b2:       220f            movel %sp,%d1
>>>>>>        59b4:       0281 ffff e000  andil #-8192,%d1
>>>>>>        59ba:       2041            moveal %d1,%a0
>>>>>>        59bc:       2228 0008       movel %a0@(8),%d1
>>>>>>        59c0:       0281 00ff ff00  andil #16776960,%d1
>>>>>>        59c6:       6600 0086       bnew 5a4e <do_040writeback1+0xe2>
>>>>>>        59ca:       7201            moveq #1,%d1
>>>>>>        59cc:       4e7b 1000       movec %d1,%sfc
>>>>>>        59d0:       4e7b 1001       movec %d1,%dfc
>>>>>>        59d4:       242e fff8       movel %fp@(-8),%d2
>>>>>>        59d8:       4e5e            unlk %fp
>>>>>>        59da:       4e75            rts
>>>>>>        59dc:       4280            clrl %d0
>>>>>>        59de:       322e 0012       movew %fp@(18),%d1
>>>>>>        59e2:       206e 000c       moveal %fp@(12),%a0
>>>>>>        59e6:       0e50 1800       movesw %d1,%a0@
>>>>>>        59ea:       220f            movel %sp,%d1
>>>>>>        59ec:       0281 ffff e000  andil #-8192,%d1
>>>>>>        59f2:       2041            moveal %d1,%a0
>>>>>>        59f4:       2228 0008       movel %a0@(8),%d1
>>>>>>        59f8:       0281 00ff ff00  andil #16776960,%d1
>>>>>>        59fe:       67ca            beqs 59ca <do_040writeback1+0x5e>
>>>>>>        5a00:       604c            bras 5a4e <do_040writeback1+0xe2>
>>>>>>        5a02:       4280            clrl %d0
>>>>>>        5a04:       222e 0010       movel %fp@(16),%d1
>>>>>>        5a08:       206e 000c       moveal %fp@(12),%a0
>>>>>>        5a0c:       0e90 1800       movesl %d1,%a0@
>>>>>>        5a10:       220f            movel %sp,%d1
>>>>>>        5a12:       0281 ffff e000  andil #-8192,%d1
>>>>>>        5a18:       2041            moveal %d1,%a0
>>>>>>        5a1a:       2228 0008       movel %a0@(8),%d1
>>>>>>        5a1e:       0281 00ff ff00  andil #16776960,%d1
>>>>>>        5a24:       67a4            beqs 59ca <do_040writeback1+0x5e>
>>>>>>        5a26:       6026            bras 5a4e <do_040writeback1+0xe2>
>>>>>>        ...
>>>>>>
>>>>>> The cause is a deliberately misaligned access in the 
>>>>>> 'bad_end_addr' test
>>>>>> case in the 'sysbadaddr' stressor. The location being accessed here,
>>>>>> 0xc043dfff, was contrived to span the boundary between a r/w 
>>>>>> anonymous page
>>>>>> and an unmapped page. The address was then passed to the getcwd 
>>>>>> syscall
>>>>>> which faulted in copy_to_user().
>>>>>>
>>>>>> The fault for the mapped page appears to be handled okay -- up until
>>>>>> do_040writeback1() called put_user() which produced a second fault 
>>>>>> due to
>>>>>> the unmapped page.
>>>>>>
>>>>>> Michael Schmitz helpfully deciphered the oops and explained the 
>>>>>> exception
>>>>>> processing leading up to it.
>>>>>>
>>>>>>        "regs->pc does point to the PC in the format 7 frame which 
>>>>>> is the PC
>>>>>>        the fault was detected at, but not (in case of a writeback 
>>>>>> fault)
>>>>>>        the PC of the faulting instruction [that is, MOVES.L].
>>>>>>
>>>>>>        "The writeback would still cross the page boundary, and 
>>>>>> fault if the
>>>>>>        unmapped page still isn't present. We would not see the PC 
>>>>>> of the
>>>>>>        movesl in that case, and fail to find the PC in the exception
>>>>>>        table."
>>>>>>
>>>>>> One solution is to add a NOP instruction after the MOVES.L to 
>>>>>> flush the
>>>>>> pipeline and take the fault. That way, the PC value in the 
>>>>>> exception frame
>>>>>> becomes dependable so the exception table works.
>>>>>>
>>>>>> Theoretically, there seems to be another bug in the existing code. If
>>>>>> the instruction following the MOVES faulted, then after the fixup,
>>>>>> execution would resume at the instruction which caused the fault. 
>>>>>> This
>>>>>> appears to be a loop. After this patch, that cannot happen.
>>>>>>
>>>>>> Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
>>>>>> Signed-off-by: Finn Thain <fthain@linux-m68k.org>
>>>>>> ---
>>>>>>     arch/m68k/include/asm/uaccess.h | 10 ++++++----
>>>>>>     1 file changed, 6 insertions(+), 4 deletions(-)
>>>>>>
>>>>>> diff --git a/arch/m68k/include/asm/uaccess.h b/arch/m68k/include/ 
>>>>>> asm/uaccess.h
>>>>>> index 31d133faa45e..728a6dfb7414 100644
>>>>>> --- a/arch/m68k/include/asm/uaccess.h
>>>>>> +++ b/arch/m68k/include/asm/uaccess.h
>>>>>> @@ -31,11 +31,12 @@
>>>>>>     #define __put_user_asm(inst, res, x, ptr, bwl, reg, err) \
>>>>>>     asm volatile ("\n"                                  \
>>>>>>         "1:     "inst"."#bwl"   %2,%1\n"                \
>>>>>> -     "2:\n"                                          \
>>>>>> +     "2:     nop\n"                                  \
>>>>>> +     "3:\n"                                          \
>>>>>>         "       .section .fixup,\"ax\"\n"               \
>>>>>>         "       .even\n"                                \
>>>>>>         "10:    moveq.l %3,%0\n"                        \
>>>>>> -     "       jra 2b\n"                               \
>>>>>> +     "       jra 3b\n"                               \
>>>>>>         "       .previous\n"                            \
>>>>>>         "\n"                                            \
>>>>>>         "       .section __ex_table,\"a\"\n"            \
>>>>>> @@ -53,11 +54,12 @@ do 
>>>>>> {                                                              \
>>>>>>         asm volatile ("\n"                                      \
>>>>>>                 "1:     "inst".l %2,(%1)+\n"                    \
>>>>>>                 "2:     "inst".l %R2,(%1)\n"                    \
>>>>>> -             "3:\n"                                          \
>>>>>> +             "3:     nop\n"                                  \
>>>>>> +             "4:\n"                                          \
>>>>>>                 "       .section .fixup,\"ax\"\n"               \
>>>>>>                 "       .even\n"                                \
>>>>>>                 "10:    movel %3,%0\n"                          \
>>>>>> -             "       jra 3b\n"                               \
>>>>>> +             "       jra 4b\n"                               \
>>>>>>                 "       .previous\n"                            \
>>>>>>                 "\n"                                            \
>>>>>>                 "       .section __ex_table,\"a\"\n"            \
>>>>>> Reviewed-by: Michael Schmitz <schmitzmic@gmail.com>
>>>>>>
>>>>>> in case it helps.
>>>>>>
>>>>>> Anecdotally, I've seen similar uaccess faults causing kernel oops 
>>>>>> on 030
>>>>>> as well, so this may not merely be an artificial unaligned access 
>>>>>> issue.
>>>> Shouldn't this include
>>>> https://lore.kernel.org/20240429030945.22451-3-schmitzmic@gmail.com
>>>> ?
>>> Along with https://lore.kernel.org/20240429030945.22451-2- 
>>> schmitzmic@gmail.com, perhaps.
>>>
>>> Can't recall the details, but Finn's solution was to add a nop in 
>>> order to make the fault PC reliable. That's not a great deal of 
>>> overhead in the case of put_user(). My changes to copy_to_user try to 
>>> avoid flushing the pipeline inside of a loop, and only use the nop 
>>> after exiting the loop. The cost of that is larger exception tables. 
>>> Not sure which is worse.
>>>
>>> What's more, I can't be certain that my 'fault is taken two 
>>> instructions past faulting instruction' observation is 100% reliable. 
>>> AFAIK this has only been tested on 030 and 040. Coldfire(MMU) and 060 
>>> tests are still missing.
>> ColdFire seems to have more, or maybe just different problems for 
>> "stress-ng --sysbadaddr -1".
>> Running a kernel with and without this patch resulted in:
>>
>> # stress-ng --sysbadaddr -1process '/usr/bin/stress-ng' started with 
>> executable stack
>> stress-ng: info:  [40] defaulting to a 86400 second (1 day, 0.00 secs) 
>> run per stressor
>> stress-ng: info:  [40] dispatching hogs: 1 sysbadaddr
>> *** ILLEGAL INSTRUCTION ***   FORMAT=4
>> Current process id is 250
>> BAD KERNEL TRAP: 00000000
>> Modules linked in:
>> PC: [<00000000>] 0x0
>> SR: 2004  SP: 013da6cc  a2: bfccd950
>> d0: 00000000    d1: 00000002    d2: bfccd814    d3: 00000001
>> d4: 80174c1e    d5: 800bd9ac    a0: 800bd9ac    a1: 800bdb54
>> Process stress-ng-sysba (pid: 250, task=f7e050bd)
>> Frame format=4 eff addr=44090004 pc=80115870
>> Stack from 013ea000:
>> Call Trace:
>> Code: 0000 0000 0000 0000 0000 0000 0000 0000 <0000> 0000 0000 0000 
>> 0000 0000 0000 0000 0002 a6c4 0002 a6c4 0002 a6c4 0002 a6c4
>> Disabling lock debugging due to kernel taint
>> note: stress-ng-sysba[250] exited with irqs disabled
>> *** ILLEGAL INSTRUCTION ***   FORMAT=4
>> Current process id is 42
>> BAD KERNEL TRAP: 00000000
>> Modules linked in:
>> PC: [<00000000>] 0x0
>> SR: 2004  SP: affae50f  a2: 012d1f80
>> d0: 00000000    d1: 0000000b    d2: 01057db0    d3: 00035b9a
>> d4: 00000000    d5: bfccd814    a0: bfccd814    a1: 0050c21c
>> Process stress-ng-sysba (pid: 42, task=1b5bfafe)
>> Frame format=4 eff addr=480a2004 pc=00035ab2
>> Stack from 013d3f28:
>>          00000000 000000fa 00000000 00000004 01057db0 00000000 
>> 0000000b 00000000
>>          00000000 012d1f80 000336b4 00000100 00000122 fffffff6 
>> bfccd76c 00035c1c
>>          000000fa bfccd814 00000000 00000000 bfccd76c 00000000 
>> 0000000a 000000fa
>>          00000000 8011685e 80119990 00935414 0002ce18 013d3fcc 
>> 00000000 00000000
>>          00000000 00000000 8011685e 013d3fcc 00000005 bfccd7a0 
>> 80010686 0002a6bc
>>          0002df3c 000000fa bfccd814 00000000 00000000 bfccd814 
>> 00000000 800bdb54
>> Call Trace: [<000336b4>] child_wait_callback+0x0/0x86
>>   [<00035c1c>] sys_wait4+0x82/0x92
>>   [<0002ce18>] buserr_c+0x130/0x1dc
>>   [<0002a6bc>] buserr+0x28/0x30
>>   [<0002df3c>] system_call+0x58/0xac
>>
>>
>> I have not debugged any further yet.
> 
> This rather looks like stack corruption - does that ColdFire system pass 
> signal tests OK?

I am not sure. I can run them and find out.
I am traveling at the moment, so don't have access to the board
right now. As soon as I get to my lab I will check.


> That said - the register dump is misleading - from what I've read, 
> ColdFire only has a single exception frame format, 2 longwords in size. 
> The first longword of that frame contains frame format code, fault 
> status, vector no. and status register. This is what's shown in the 'eff 
> addr' longword here. The following longword is the actual faulting 
> instruction PC as saved in the exception frame by the processor - this 
> does not match what is given as PC at the start of the register dump 
> (saved by the kernel on exceptio processing entry).

I noticed that. It all looks a bit inconsistent.


> I don't know what kind of ColdFire processor you used, so can't work out 
> what happened on exception entry and in particular, where the PC in the 
> kernel stack frame was taken from. No matter what, I think we ought to 
> clean up the frame format 4 section in show_registers() to account for 
> the different frame format on ColdFire, then perhaps also use the 
> correct PC to dump the infringing code section. Makes no sense to try 
> and debug this without correct debug info ...
> 
> I can prepare a patch to show what I mean, but can neither compile or 
> test that...

I can try anything you want - at least once I get back to the lab.

FWIW I am testing on a M5475 processor (on a M5475-EVB board).
That is the m5475evb_defconfig in mainline kernels. So if you want to
compile check any code changes just build for that.

Regards
Greg



  reply	other threads:[~2026-10-08 13:41 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2026-10-01  9:08 Finn Thain
2026-10-01 18:57 ` Michael Schmitz
2026-10-02 16:28   ` Geert Uytterhoeven
2026-10-02 23:55     ` Michael Schmitz
2026-10-03  1:23       ` Daniel Palmer
2026-10-05 12:42       ` Greg Ungerer
2026-10-07 23:54         ` Michael Schmitz
2026-10-08 13:41           ` Greg Ungerer [this message]
2026-10-09  6:56             ` Michael Schmitz
2026-10-12 13:23           ` Greg Ungerer
2026-10-12 17:42             ` Michael Schmitz
2026-10-03  5:30     ` Finn Thain

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=92bfc41e-3c37-4485-b633-f41f58f0b0bd@linux-m68k.org \
    --to=gerg@linux-m68k.org \
    --cc=fthain@linux-m68k.org \
    --cc=geert@linux-m68k.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-m68k@lists.linux-m68k.org \
    --cc=schmitzmic@gmail.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox

all inboxes | Powered by JetHome®