From: Joao Moreira <jmoreira@suse.de>
To: Miroslav Benes <mbenes@suse.cz>, Joe Lawrence <joe.lawrence@redhat.com>
Cc: live-patching@vger.kernel.org, pmladek@suse.cz, jikos@suse.cz,
nstange@suse.de, jpoimboe@redhat.com, khlebnikov@yandex-team.ru,
jeyu@kernel.org, matz@suse.de, linux-kernel@vger.kernel.org,
yamada.masahiro@socionext.com, linux-kbuild@vger.kernel.org,
michal.lkml@markovi.net
Subject: Re: [PATCH v2 3/8] livepatch: Add klp-convert tool
Date: Tue, 26 Mar 2019 17:13:57 -0300 [thread overview]
Message-ID: <04f6c9c4-88c6-41b3-fbf8-2409b3822e0b@suse.de> (raw)
In-Reply-To: <alpine.LSU.2.21.1903202021570.29796@pobox.suse.cz>
On 3/20/19 4:36 PM, Miroslav Benes wrote:
>>> [ ... snip ... ]
>>> +
>>> +/* Checks if sympos is valid, otherwise prints valid sympos list */
>>> +static bool valid_sympos(struct sympos *sp)
>>> +{
>>> + struct symbol_entry *e;
>>> + int counter = 0;
>>> +
>>> + list_for_each_entry(e, &symbols, list) {
>>> + if ((strcmp(e->symbol_name, sp->symbol_name) == 0) &&
>>> + (strcmp(e->object_name, sp->object_name) == 0)) {
>>> + if (counter == sp->pos)
>>> + return true;
>>> + counter++;
>>> + }
>>> + }
>>> +
>>> + WARN("Provided KLP_SYMPOS does not match a symbol: %s.%s,%d",
>>> + sp->object_name, sp->symbol_name, sp->pos);
>>> + print_valid_module_relocs(sp->symbol_name);
>>> +
>>> + return false;
>>> +}
>>
>> I believe there is an off-by-one error condition either here, or in
>> livepatch kernel core sympos disambiguator code (in which case, external
>> tools like kpatch-build would also need to be adjusted).
>>
>> The scenarios that I've observed using klp-convert and kpatch-build:
>>
>> sympos == 0, uniquely named symbol
>> sympos == 1, non-unique symbol name, first instance
>> sympos == 2, non-unique symbol name, second instance
>> ...
>
> Yes, that is exactly how we defined it back then.
>
>> When I built a klp-convert selftest, I made sure that the target module
>> contained multiple symbols of the same name across two .o files.
>> However, when I tried to use a KLP_MODULE_RELOC annotation in the
>> livepatch to resolve to the second (ie, last) symbol, using a sympos=2,
>> klp-convert kept telling me that the only valid KLP_SYMPOS were 0 and 1.
>>
>> The following code adjusts klp-convert to match the sympos logic, as I
>> understand it in livepatch/core.c and kpatch-build;
>>
>> [squash] livepatch/klp-convert: fix klp-convert off-by-one sympos
>> https://github.com/torvalds/linux/commit/1ed8e5baa98f7920bbbaa562278b3ed44552e01f
>>
>> This change also adds a check that sympos == 0 is in fact a uniquely
>> named symbol.
>
Nice catch, thanks for it. All looks good to me.
> Looks good to me. Maybe we can even make it simpler and use similar
> approach as in klp_find_callback() and klp_find_object_symbol() from
> kernel/livepatch/core.c. On the other hand, given that we want to print
> useful output in all cases, the code might happen to be more complicated
> and definitely ugly.
I don't have an opinion here.
>
>>> [ ... snip ... ]
>>> +
>>> +int main(int argc, const char **argv)
>>> +{
>>> + const char *klp_in_module, *klp_out_module, *symbols_list;
>>> + struct rela *rela, *tmprela;
>>> + struct section *sec, *aux;
>>> + struct sympos sp;
>>> + struct elf *klp_elf;
>>> +
>>> + if (argc != 4) {
>>> + WARN("Usage: %s <Symbols.list> <input.ko> <out.ko>", argv[0]);
>>> + return -1;
>>> + }
>>> +
>>> + symbols_list = argv[1];
>>> + klp_in_module = argv[2];
>>> + klp_out_module = argv[3];
>>> +
>>> + klp_elf = elf_open(klp_in_module);
>>> + if (!klp_elf) {
>>> + WARN("Unable to read elf file %s\n", klp_in_module);
>>> + return -1;
>>> + }
>>> +
>>> + if (!load_syms_lists(symbols_list))
>>> + return -1;
>>> +
>>> + if (!load_usr_symbols(klp_elf)) {
>>> + WARN("Unable to load user-provided sympos");
>>> + return -1;
>>> + }
>>> +
>>> + list_for_each_entry_safe(sec, aux, &klp_elf->sections, list) {
>>> + if (!is_rela_section(sec))
>>> + continue;
>>> +
>>> + list_for_each_entry_safe(rela, tmprela, &sec->relas, list) {
>>> + if (!must_convert(rela->sym))
>>> + continue;
>>> +
>>> + if (!find_missing_position(rela->sym, &sp)) {
>>> + WARN("Unable to find missing symbol");
>>
>> A really minor suggestion, but I found it useful to tell the user
>> exactly which symbol was missing:
>>
>> [squash] livepatch/klp-convert: add missing symbol name to warning
>> https://github.com/torvalds/linux/commit/7a41a8f3d9b9e11f0c75914cb925af1486c1ecc6
>
> I think that Joao has exactly the same hunk staged somewhere
> already. You are right. The message is not much informative.
>
Yes, someone had the same idea for the klp-convert which we use inside
SUSE. But it did not make it to this patch-set.
Nice that Joe added it here :)
> Miroslav
>
next prev parent reply other threads:[~2019-03-26 20:14 UTC|newest]
Thread overview: 30+ messages / expand[flat|nested] mbox.gz Atom feed top
[not found] <20190301141313.15057-1-jmoreira@suse.de>
2019-03-18 19:18 ` [PATCH v2 0/8] klp-convert Joe Lawrence
2019-03-26 20:18 ` Joao Moreira
2019-03-26 21:03 ` Joe Lawrence
2019-04-04 11:49 ` Miroslav Benes
2019-04-04 13:19 ` Joe Lawrence
[not found] ` <20190301141313.15057-3-jmoreira@suse.de>
2019-03-18 19:19 ` [PATCH v2 2/8] kbuild: Support for Symbols.list creation Joe Lawrence
2019-03-20 19:08 ` Miroslav Benes
2019-03-26 14:40 ` Joao Moreira
2019-03-26 16:15 ` Joe Lawrence
2019-03-26 18:13 ` Joao Moreira
2019-03-26 20:53 ` Joe Lawrence
2019-03-28 20:17 ` Joe Lawrence
2019-04-01 19:35 ` Joe Lawrence
2019-04-03 12:48 ` Miroslav Benes
2019-04-03 19:10 ` Joe Lawrence
2019-04-04 9:14 ` Miroslav Benes
2019-04-04 10:59 ` Miroslav Benes
[not found] ` <20190301141313.15057-4-jmoreira@suse.de>
2019-03-18 19:20 ` [PATCH v2 3/8] livepatch: Add klp-convert tool Joe Lawrence
2019-03-20 19:36 ` Miroslav Benes
2019-03-26 20:13 ` Joao Moreira [this message]
[not found] ` <20190301141313.15057-7-jmoreira@suse.de>
2019-03-18 19:20 ` [PATCH v2 6/8] modpost: Add modinfo flag to livepatch modules Joe Lawrence
[not found] ` <20190301141313.15057-8-jmoreira@suse.de>
2019-03-18 19:21 ` [PATCH v2 7/8] livepatch: Add sample livepatch module Joe Lawrence
[not found] ` <20190301141313.15057-9-jmoreira@suse.de>
2019-03-18 19:21 ` [PATCH v2 8/8] documentation: Update on livepatch elf format Joe Lawrence
2019-03-20 19:58 ` Miroslav Benes
[not found] ` <20190301141313.15057-6-jmoreira@suse.de>
2019-03-18 19:20 ` [PATCH v2 5/8] modpost: Integrate klp-convert Joe Lawrence
2019-03-22 14:54 ` Joe Lawrence
2019-03-22 16:37 ` Joao Moreira
2019-03-22 18:29 ` Joe Lawrence
2019-04-04 11:31 ` Miroslav Benes
2019-04-04 13:55 ` Joao Moreira
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=04f6c9c4-88c6-41b3-fbf8-2409b3822e0b@suse.de \
--to=jmoreira@suse.de \
--cc=jeyu@kernel.org \
--cc=jikos@suse.cz \
--cc=joe.lawrence@redhat.com \
--cc=jpoimboe@redhat.com \
--cc=khlebnikov@yandex-team.ru \
--cc=linux-kbuild@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=live-patching@vger.kernel.org \
--cc=matz@suse.de \
--cc=mbenes@suse.cz \
--cc=michal.lkml@markovi.net \
--cc=nstange@suse.de \
--cc=pmladek@suse.cz \
--cc=yamada.masahiro@socionext.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).