From: Christophe Leroy <christophe.leroy@csgroup.eu>
To: "Naveen N. Rao" <naveen.n.rao@linux.vnet.ibm.com>,
"linuxppc-dev@lists.ozlabs.org" <linuxppc-dev@lists.ozlabs.org>,
Sathvika Vasireddy <sv@linux.ibm.com>,
Sathvika Vasireddy <sv@linux.vnet.ibm.com>
Cc: Marc Zyngier <maz@kernel.org>, "aik@ozlabs.ru" <aik@ozlabs.ru>,
"linux-kernel@vger.kernel.org" <linux-kernel@vger.kernel.org>,
"rostedt@goodmis.org" <rostedt@goodmis.org>,
"peterz@infradead.org" <peterz@infradead.org>,
"mingo@redhat.com" <mingo@redhat.com>,
"paulus@samba.org" <paulus@samba.org>,
"jpoimboe@redhat.com" <jpoimboe@redhat.com>,
"mbenes@suse.cz" <mbenes@suse.cz>,
Chen Zhongjin <chenzhongjin@huawei.com>,
Linux ARM <linux-arm-kernel@lists.infradead.org>
Subject: Re: [RFC PATCH v3 11/12] powerpc: Remove unreachable() from WARN_ON()
Date: Thu, 30 Jun 2022 09:58:48 +0000 [thread overview]
Message-ID: <da86c612-186d-364f-cc36-bcf942a97083@csgroup.eu> (raw)
In-Reply-To: <1656572413.pbaqjnrrcl.naveen@linux.ibm.com>
Le 30/06/2022 à 10:05, Naveen N. Rao a écrit :
> Christophe Leroy wrote:
>> Hi Sathvika,
>>
>> Adding ARM people as they seem to face the same kind of problem (see
>> https://patchwork.kernel.org/project/linux-kbuild/patch/20220623014917.199563-33-chenzhongjin@huawei.com/)
>>
>>
>> Le 27/06/2022 à 17:35, Sathvika Vasireddy a écrit :
>>>
>>> On 25/06/22 12:16, Christophe Leroy wrote:
>>>>
>>>> Le 24/06/2022 à 20:32, Sathvika Vasireddy a écrit :
>>>>> objtool is throwing *unannotated intra-function call*
>>>>> warnings with a few instructions that are marked
>>>>> unreachable. Remove unreachable() from WARN_ON()
>>>>> to fix these warnings, as the codegen remains same
>>>>> with and without unreachable() in WARN_ON().
>>>> Did you try the two exemples described in commit 1e688dd2a3d6
>>>> ("powerpc/bug: Provide better flexibility to WARN_ON/__WARN_FLAGS()
>>>> with
>>>> asm goto") ?
>>>>
>>>> Without your patch:
>>>>
>>>> 00000640 <test>:
>>>> 640: 81 23 00 84 lwz r9,132(r3)
>>>> 644: 71 29 40 00 andi. r9,r9,16384
>>>> 648: 40 82 00 0c bne 654 <test+0x14>
>>>> 64c: 80 63 00 0c lwz r3,12(r3)
>>>> 650: 4e 80 00 20 blr
>>>> 654: 0f e0 00 00 twui r0,0
>>>>
>>>> 00000658 <test9w>:
>>>> 658: 2c 04 00 00 cmpwi r4,0
>>>> 65c: 41 82 00 0c beq 668 <test9w+0x10>
>>>> 660: 7c 63 23 96 divwu r3,r3,r4
>>>> 664: 4e 80 00 20 blr
>>>> 668: 0f e0 00 00 twui r0,0
>>>> 66c: 38 60 00 00 li r3,0
>>>> 670: 4e 80 00 20 blr
>>>>
>>>>
>>>> With your patch:
>>>>
>>>> 00000640 <test>:
>>>> 640: 81 23 00 84 lwz r9,132(r3)
>>>> 644: 71 29 40 00 andi. r9,r9,16384
>>>> 648: 40 82 00 0c bne 654 <test+0x14>
>>>> 64c: 80 63 00 0c lwz r3,12(r3)
>>>> 650: 4e 80 00 20 blr
>>>> 654: 0f e0 00 00 twui r0,0
>>>> 658: 4b ff ff f4 b 64c <test+0xc> <==
>>>>
>>>> 0000065c <test9w>:
>>>> 65c: 2c 04 00 00 cmpwi r4,0
>>>> 660: 41 82 00 0c beq 66c <test9w+0x10>
>>>> 664: 7c 63 23 96 divwu r3,r3,r4
>>>> 668: 4e 80 00 20 blr
>>>> 66c: 0f e0 00 00 twui r0,0
>>>> 670: 38 60 00 00 li r3,0 <==
>>>> 674: 4e 80 00 20 blr <==
>>>> 678: 38 60 00 00 li r3,0
>>>> 67c: 4e 80 00 20 blr
>>>>
>>> The builtin variant of unreachable (__builtin_unreachable()) works.
>>>
>>> How about using that instead of unreachable() ?
>>>
>>>
>>
>> In fact the problem comes from the macro annotate_unreachable() which
>> is called by unreachable() before calling __build_unreachable().
>>
>> Seems like this macro adds (after the unconditional trap twui) a call
>> to an empty function whose address is listed in section
>> .discard.unreachable
>>
>> 1c78: 00 00 e0 0f twui r0,0
>> 1c7c: 55 e7 ff 4b bl 3d0
>> <qdisc_root_sleeping_lock.part.0>
>>
>>
>> RELOCATION RECORDS FOR [.discard.unreachable]:
>> OFFSET TYPE VALUE
>> 0000000000000000 R_PPC64_REL32 .text+0x00000000000003d0
>>
>> The problem is that that function has size 0:
>>
>> 00000000000003d0 l F .text 0000000000000000
>> qdisc_root_sleeping_lock.part.0
>>
>>
>> And objtool is not prepared for a function with size 0.
>
> annotate_unreachable() seems to have been introduced in commit
> 649ea4d5a624f0 ("objtool: Assume unannotated UD2 instructions are dead
> ends").
>
> Objtool considers 'ud2' instruction to be fatal, so BUG() has
> __builtin_unreachable(), rather than unreachable(). See commit
> bfb1a7c91fb775 ("x86/bug: Merge annotate_reachable() into _BUG_FLAGS()
> asm"). For the same reason, __WARN_FLAGS() is annotated with
> _ASM_REACHABLE so that objtool can differentiate warnings from a BUG().
>
> On powerpc, we use trap variants for both and don't have a special
> instruction for a BUG(). As such, for _WARN_FLAGS(), using
> __builtin_unreachable() suffices to achieve optimal code generation from
> the compiler. Objtool would consider subsequent instructions to be
> reachable. For BUG(), we can continue to use unreachable() so that
> objtool can differentiate these from traps used in warnings.
Not sure I understand what you mean.
__WARN_FLAGS() and BUG() both use 'twui' which is unconditionnal trap,
as such both are the same.
On the other side, WARN_ON() and BUG_ON() use tlbnei which is a
conditionnel trap.
>
>>
>> The following changes to objtool seem to fix the problem, most warning
>> are gone with that change.
>>
>> diff --git a/tools/objtool/elf.c b/tools/objtool/elf.c
>> index 63218f5799c2..37c0a268b7ea 100644
>> --- a/tools/objtool/elf.c
>> +++ b/tools/objtool/elf.c
>> @@ -77,6 +77,8 @@ static int symbol_by_offset(const void *key, const
>> struct rb_node *node)
>>
>> if (*o < s->offset)
>> return -1;
>> + if (*o == s->offset && !s->len)
>> + return 0;
>> if (*o >= s->offset + s->len)
>> return 1;
>>
>> @@ -400,7 +402,7 @@ static void elf_add_symbol(struct elf *elf, struct
>> symbol *sym)
>> * Don't store empty STT_NOTYPE symbols in the rbtree. They
>> * can exist within a function, confusing the sorting.
>> */
>> - if (!sym->len)
>> + if (sym->type == STT_NOTYPE && !sym->len)
>> rb_erase(&sym->node, &sym->sec->symbol_tree);
>> }
>
> Is there a reason to do this, rather than change __WARN_FLAGS() to use
> __builtin_unreachable()? Or, are you seeing an issue with unreachable()
> elsewhere in the kernel?
>
At the moment I'm trying to understand what the issue is, and explore
possible fixes. I guess if we tell objtool that after 'twui' subsequent
instructions are unreachable, then __builtin_unreachable() is enough.
I think we should also understand why annotate_unreachable() gives us a
so bad result and see if it can be changed to something cleaner than a
'bl' to an empty function that has no instructions.
next prev parent reply other threads:[~2022-06-30 9:59 UTC|newest]
Thread overview: 41+ messages / expand[flat|nested] mbox.gz Atom feed top
2022-06-24 18:32 [RFC PATCH v3 00/12] objtool: Enable and implement --mcount option on powerpc Sathvika Vasireddy
2022-06-24 18:32 ` [RFC PATCH v3 01/12] objtool: Fix SEGFAULT Sathvika Vasireddy
2022-07-08 15:10 ` Christophe Leroy
2022-06-24 18:32 ` [RFC PATCH v3 02/12] objtool: Use target file endianness instead of a compiled constant Sathvika Vasireddy
2022-06-24 18:32 ` [RFC PATCH v3 03/12] objtool: Use target file class size " Sathvika Vasireddy
2022-07-08 17:35 ` Christophe Leroy
2022-06-24 18:32 ` [RFC PATCH v3 04/12] objtool: Add --mnop as an option to --mcount Sathvika Vasireddy
2022-06-24 18:32 ` [RFC PATCH v3 05/12] powerpc: Skip objtool from running on VDSO files Sathvika Vasireddy
2022-06-24 18:32 ` [RFC PATCH v3 06/12] objtool: Read special sections with alts only when specific options are selected Sathvika Vasireddy
2022-06-24 18:32 ` [RFC PATCH v3 07/12] objtool: Use macros to define arch specific reloc types Sathvika Vasireddy
2022-07-04 11:14 ` Peter Zijlstra
2022-07-04 15:53 ` Christophe Leroy
2022-07-04 16:18 ` Peter Zijlstra
2022-06-24 18:32 ` [RFC PATCH v3 08/12] objtool: Add arch specific function arch_ftrace_match() Sathvika Vasireddy
2022-06-24 18:32 ` [RFC PATCH v3 09/12] objtool/powerpc: Enable objtool to be built on ppc Sathvika Vasireddy
2022-06-24 18:32 ` [RFC PATCH v3 10/12] objtool/powerpc: Add --mcount specific implementation Sathvika Vasireddy
2022-06-24 18:32 ` [RFC PATCH v3 11/12] powerpc: Remove unreachable() from WARN_ON() Sathvika Vasireddy
2022-06-25 6:46 ` Christophe Leroy
2022-06-27 15:21 ` Sathvika Vasireddy
2022-06-27 15:35 ` Sathvika Vasireddy
2022-06-27 15:46 ` Christophe Leroy
2022-06-29 18:30 ` Christophe Leroy
2022-06-30 8:05 ` Naveen N. Rao
2022-06-30 9:58 ` Christophe Leroy [this message]
2022-06-30 10:33 ` Christophe Leroy
2022-06-30 10:37 ` Naveen N. Rao
2022-06-30 15:58 ` Segher Boessenkool
2022-07-04 12:01 ` Peter Zijlstra
2022-07-04 11:43 ` Peter Zijlstra
2022-07-01 2:13 ` Chen Zhongjin
2022-07-01 6:56 ` Sathvika Vasireddy
2022-07-01 11:40 ` [RFC PATCH v3 11/12] powerpc: Remove unreachable() from WARN_ON() (gcc issue ?) Christophe Leroy
2022-07-04 11:45 ` [RFC PATCH v3 11/12] powerpc: Remove unreachable() from WARN_ON() Peter Zijlstra
2022-07-04 12:34 ` Christophe Leroy
2022-07-05 15:48 ` Segher Boessenkool
2022-07-04 12:05 ` Peter Zijlstra
2022-07-04 12:44 ` Christophe Leroy
2022-07-04 14:19 ` Peter Zijlstra
2022-06-24 18:32 ` [RFC PATCH v3 12/12] objtool/powerpc: Fix unannotated intra-function call warnings Sathvika Vasireddy
2022-07-08 15:06 ` [RFC PATCH v3 00/12] objtool: Enable and implement --mcount option on powerpc Christophe Leroy
2022-07-08 15:42 ` Christophe Leroy
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=da86c612-186d-364f-cc36-bcf942a97083@csgroup.eu \
--to=christophe.leroy@csgroup.eu \
--cc=aik@ozlabs.ru \
--cc=chenzhongjin@huawei.com \
--cc=jpoimboe@redhat.com \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linuxppc-dev@lists.ozlabs.org \
--cc=maz@kernel.org \
--cc=mbenes@suse.cz \
--cc=mingo@redhat.com \
--cc=naveen.n.rao@linux.vnet.ibm.com \
--cc=paulus@samba.org \
--cc=peterz@infradead.org \
--cc=rostedt@goodmis.org \
--cc=sv@linux.ibm.com \
--cc=sv@linux.vnet.ibm.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;
as well as URLs for NNTP newsgroup(s).