linux-kernel - Re: [PATCH] objtool: Support multiple rodata sections.

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-ID: <20180726163647.xxn5u6c2ekxtk2b4@treble>
Date:   Thu, 26 Jul 2018 11:36:47 -0500
From:   Josh Poimboeuf <jpoimboe@...hat.com>
To:     Allan Xavier <allan.x.xavier@...cle.com>
Cc:     Peter Zijlstra <peterz@...radead.org>, linux-kernel@...r.kernel.org
Subject: Re: [PATCH] objtool: Support multiple rodata sections.

On Wed, Jul 25, 2018 at 03:17:53PM +0100, Allan Xavier wrote:
> This commit adds support for processing switch jump tables in objects
> with multiple .rodata sections, such as those created when using
> -ffunction-sections and -fdata-sections.  Currently, objtool always
> looks in .rodata for jump table information which results in many
> "sibling call from callable instruction with modified stack frame"
> warnings with objects compiled using those flags.

Thanks for the patch!  kpatch-build also uses those flags and has a
similar issue, so it's good this will be useful for multiple users.

> @@ -930,10 +931,13 @@ static struct rela *find_switch_table(struct objtool_file *file,
>  		/* look for a relocation which references .rodata */
>  		text_rela = find_rela_by_dest_range(insn->sec, insn->offset,
>  						    insn->len);
> -		if (!text_rela || text_rela->sym != file->rodata->sym)
> +		if (!text_rela || !text_rela->sym->sec->rodata
> +			       || text_rela->sym != text_rela->sym->sec->sym)
>  			continue;

This last condition is a bit hard to follow.  I think we can just check
that the sym type is STT_SECTION.  Also the formatting deviates a bit
from the kernel style.  How about something like this?

		if (!text_rela || text_rela->sym->type != STT_SECTION ||
		    !text_rela->sym->sec->rodata)
			continue;

> @@ -2133,6 +2137,20 @@ static void cleanup(struct objtool_file *file)
>  	elf_close(file->elf);
>  }
>  
> +static bool mark_rodata(struct objtool_file *file)
> +{
> +	struct section *sec;
> +	bool found = false;
> +
> +	for_each_sec(file, sec)

Please use brackets for the outer loop.

> +		if (strstr(sec->name, ".rodata") == sec->name) {
> +			sec->rodata = true;
> +			found = true;

This check is too broad.  It will also match the following sections:

.rodata.str1.8
.rodata.str1.1
.rela.rodata
.init.rodata

Also, the loop should break once it finds an rodata section.

> +		}
> +
> +	return found;
> +}
> +
>  int check(const char *_objname, bool orc)
>  {
>  	struct objtool_file file;
> @@ -2147,10 +2165,10 @@ int check(const char *_objname, bool orc)
>  	INIT_LIST_HEAD(&file.insn_list);
>  	hash_init(file.insn_hash);
>  	file.whitelist = find_section_by_name(file.elf, ".discard.func_stack_frame_non_standard");
> -	file.rodata = find_section_by_name(file.elf, ".rodata");
>  	file.c_file = find_section_by_name(file.elf, ".comment");
>  	file.ignore_unreachables = no_unreachable;
>  	file.hints = false;
> +	file.rodata = mark_rodata(&file);

I believe it would be cleaner to call mark_rodata() from
decode_sections().  And file.rodata bool should be set from within
mark_rodata().

-- 
Josh