mirror of https://lore.kernel.org/lkml/
 help / color / mirror / Atom feed
From: Guenter Roeck <linux@roeck-us.net>
To: John David Anglin <dave.anglin@bell.net>,
	David Laight <David.Laight@ACULAB.COM>,
	Charlie Jenkins <charlie@rivosinc.com>,
	Palmer Dabbelt <palmer@dabbelt.com>,
	Andrew Morton <akpm@linux-foundation.org>,
	Helge Deller <deller@gmx.de>,
	"James E.J. Bottomley" <James.Bottomley@hansenpartnership.com>,
	Parisc List <linux-parisc@vger.kernel.org>,
	Al Viro <viro@zeniv.linux.org.uk>
Cc: "linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>
Subject: Re: [PATCH v8 2/2] lib: checksum: Use aligned accesses for ip_fast_csum and csum_ipv6_magic tests
Date: Thu, 15 Feb 2024 09:06:14 -0800	[thread overview]
Message-ID: <1e302d8f-4e94-4278-b556-b8fc54956efb@roeck-us.net> (raw)
In-Reply-To: <4723822c-2acf-4c41-899c-1e3d5659d1d8@bell.net>

On 2/15/24 08:51, John David Anglin wrote:
> On 2024-02-15 10:44 a.m., Guenter Roeck wrote:
>> On 2/15/24 02:27, David Laight wrote:
>>> ...
>>>> It would be worthwhile tracking this down since there are
>>>> lots of unaligned data accesses (8-byte accesses on 4-byte aligned addresses)
>>>> when running the kernel in 64-bit mode.
>>>
>>> Hmmm....
>>> For performance reasons you really don't want any of them.
>>> The misaligned 64bit fields need an __attribute((aligned(4)) marker.
>>>
>>> If the checksum code can do them it really needs to detect
>>> and handle the misalignment.
>>>
>>> The misaligned trap handler probably ought to contain a
>>> warn_on_once() to dump stack on the first such error.
>>> They can then be fixed one at a time.
>>>
>>
>> Unaligned LDD at unwind_once+0x4a8/0x5e0
>>
>> Decoded:
>>
>> Unaligned LDD at unwind_once (arch/parisc/kernel/unwind.c:212 arch/parisc/kernel/unwind.c:243 arch/parisc/kernel/unwind.c:371 arch/parisc/kernel/unwind.c:445)
>>
>> Source:
>>
>> static bool pc_is_kernel_fn(unsigned long pc, void *fn)
>> {
>>         return (unsigned long)dereference_kernel_function_descriptor(fn) == pc;
> This looks wrong to me.  Function descriptors should always be 8-byte aligned.  I think this
> routine should return false if fn isn't 8-byte aligned.

Below you state "Code entry points only need 4-byte alignment."

I think that contradicts each other. Also, the calling code is,
for example,
	pc_is_kernel_fn(pc, syscall_exit)

I fail to see how this can be consolidated if it is ok
that syscall_exit is 4-byte aligned but, at the same time,
must be 8-byte aligned to be considered to be a kernel function.

Guenter

>> }
>>
>> Disassembled:
>>
>>  c38:   50 3c 00 00     ldd 0(r1),ret0
>>  c3c:   08 1b 02 44     copy dp,r4
>>  c40:   0f 80 10 da     ldd 0(ret0),r26    <--
>>  c44:   37 dd 3f a1     ldo -30(sp),ret1
>>  c48:   51 02 00 20     ldd 10(r8),rp
>>  c4c:   e8 40 f0 00     bve,l (rp),rp
>>  c50:   51 1b 00 30     ldd 18(r8),dp
>>
>> Hmm, interesting. See below for a possible fix. Should I submit a patch ?
>>
>> Thanks,
>> Guenter
>>
>> ---
>> diff --git a/arch/parisc/kernel/entry.S b/arch/parisc/kernel/entry.S
>> index ab23e61a6f01..1d2aab619466 100644
>> --- a/arch/parisc/kernel/entry.S
>> +++ b/arch/parisc/kernel/entry.S
>> @@ -772,6 +772,7 @@ ENTRY_CFI(_switch_to)
>>         bv      %r0(%r2)
>>         mtctl   %r25,%cr30
>>
>> +       .align  8
> Code entry points only need 4-byte alignment.
>>  ENTRY(_switch_to_ret)
>>         mtctl   %r0, %cr0               /* Needed for single stepping */
>>         callee_rest
>> @@ -1702,6 +1703,7 @@ ENTRY_CFI(sys_rt_sigreturn_wrapper)
>>         LDREG   PT_GR28(%r1),%r28  /* reload original r28 for syscall_exit */
>>  ENDPROC_CFI(sys_rt_sigreturn_wrapper)
>>
>> +       .align 8
> Ditto.
>>  ENTRY(syscall_exit)
>>         /* NOTE: Not all syscalls exit this way.  rt_sigreturn will exit
>>          * via syscall_exit_rfi if the signal was received while the process
> 
> Dave
> 


  reply	other threads:[~2024-02-15 17:06 UTC|newest]

Thread overview: 38+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-02-14 21:41 [PATCH v8 0/2] lib: checksum: Fix issues with checksum tests Charlie Jenkins
2024-02-14 21:41 ` [PATCH v8 1/2] lib: checksum: Fix type casting in checksum kunits Charlie Jenkins
2024-02-14 21:41 ` [PATCH v8 2/2] lib: checksum: Use aligned accesses for ip_fast_csum and csum_ipv6_magic tests Charlie Jenkins
2024-02-14 23:03   ` Guenter Roeck
2024-02-15  1:30     ` Charlie Jenkins
2024-02-15  1:58       ` Guenter Roeck
2024-02-15  3:00         ` John David Anglin
2024-02-15  3:35           ` Charlie Jenkins
2024-02-15  4:11             ` Charlie Jenkins
2024-02-15  8:56             ` Guenter Roeck
2024-02-15 15:36               ` Charlie Jenkins
2024-02-15 16:30                 ` Guenter Roeck
2024-02-15 16:51                   ` Charlie Jenkins
2024-02-15 17:13                     ` John David Anglin
2024-02-15 17:29                     ` Guenter Roeck
2024-02-15 15:22           ` Guenter Roeck
2024-02-16  5:54         ` Helge Deller
2024-02-16  5:25           ` Guenter Roeck
2024-02-16  7:31             ` Helge Deller
2024-02-16  6:58               ` Guenter Roeck
2024-02-16  6:09           ` Guenter Roeck
2024-02-16  6:13             ` Charlie Jenkins
2024-02-16 12:05               ` Helge Deller
2024-02-16 18:22           ` John David Anglin
2024-02-15 10:27     ` David Laight
2024-02-15 15:44       ` Guenter Roeck
2024-02-15 16:51         ` John David Anglin
2024-02-15 17:06           ` Guenter Roeck [this message]
2024-02-15 17:25             ` John David Anglin
2024-02-15 18:17               ` Guenter Roeck
2024-02-15 18:56                 ` John David Anglin
2024-02-15 21:00                   ` Guenter Roeck
2024-02-15 21:08                     ` John David Anglin
2024-02-15 18:42               ` Guenter Roeck
2024-02-15 19:41                 ` David Laight
2024-02-15 19:42                 ` Charlie Jenkins
2024-02-16  5:00                   ` Guenter Roeck
2024-02-15  1:24   ` Guenter Roeck

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=1e302d8f-4e94-4278-b556-b8fc54956efb@roeck-us.net \
    --to=linux@roeck-us.net \
    --cc=David.Laight@ACULAB.COM \
    --cc=James.Bottomley@hansenpartnership.com \
    --cc=akpm@linux-foundation.org \
    --cc=charlie@rivosinc.com \
    --cc=dave.anglin@bell.net \
    --cc=deller@gmx.de \
    --cc=linux-kernel@vger.kernel.org \
    --cc=linux-parisc@vger.kernel.org \
    --cc=palmer@dabbelt.com \
    --cc=viro@zeniv.linux.org.uk \
    /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®