mailing list of musl libc
 help / color / mirror / code / Atom feed
From: Andy Caldwell <andycaldwell@microsoft.com>
To: "musl@lists.openwall.com" <musl@lists.openwall.com>
Subject: [musl] RE: [EXTERNAL] Re: [musl] [PATCH] fix avoidable segfault in catclose
Date: Thu, 25 Jan 2024 20:10:30 +0000	[thread overview]
Message-ID: <AS4PR83MB05462D2FE519A65787A596DACB7A2@AS4PR83MB0546.EURPRD83.prod.outlook.com> (raw)
In-Reply-To: <ZbJsGMwsqlU_DY8g@voyager>

> And it has been musl policy to crash on invalid args since the beginning.

The current implementation doesn't (necessarily) crash/trap on an invalid argument, instead it invokes (C-language spec-defined) UB itself (it dereferences `(uint32_t*)((char*)cat) + 8)`, which, in the case of the `-1` handle is the address 0x7, which in turn, not being a valid address, is UB to dereference).  If you're lucky (or are compiling without optimizations/inlining) the compiler will emit a MOV that will trigger an access violation and hence a SEGV, if you're unlucky the compiler will make wild assumptions about the value of the variable passed as the arg (and for example in your first code snippet, simply delete the `if` statement, meaning `use_cat` gets called even when `catopen` fails potentially corrupting user data/state).  Crashing loudly (which requires _not_ invoking UB) on known bad inputs (a test against `-1` isn't exactly expensive) feels like it meets the "musl policy" better than the current code.

A

-----Original Message-----
From: Markus Wichmann <nullplan@gmx.net> 
Sent: Thursday, January 25, 2024 2:12 PM
To: musl@lists.openwall.com
Cc: Rich Felker <dalias@libc.org>; Ismael Luceno <ismael@iodev.co.uk>
Subject: [EXTERNAL] Re: [musl] [PATCH] fix avoidable segfault in catclose

Am Thu, Jan 25, 2024 at 08:09:49AM +0100 schrieb Ismael Luceno:
> catclose may be called with an invalid argument, particularly -1 may 
> be returned by catopen if there's an error.
>

May it, though? My copy of POSIX does not say so. Whenever a function description does not say that you can call a function with invalid arguments, you cannot do so. And it has been musl policy to crash on invalid args since the beginning.

The problem you describe sounds like your app has control flow being
approximately:

nl_catd cat = catopen(...);
if (cat != (nl_catd)-1) {
    use_cat(cat);
}
catclose(cat);

and that is just wrong control flow and can be remedied by just moving one line:

nl_catd cat = catopen(...);
if (cat != (nl_catd)-1) {
    use_cat(cat);
    catclose(cat);
}

BTW, POSIX does not say catclose() is required (or even allowed) to accept (nl_catd)-1 as argument, its description of the return value of
catopen() also says that it is only suitable for use with catclose() when successful.

Ciao,
Markus

  parent reply	other threads:[~2024-01-25 20:10 UTC|newest]

Thread overview: 17+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2024-01-25  7:09 Ismael Luceno
2024-01-25 14:05 ` Rich Felker
2024-01-25 15:28   ` Ismael Luceno
2024-01-25 15:56     ` Rich Felker
2024-01-25 14:11 ` Markus Wichmann
2024-01-25 15:30   ` Ismael Luceno
2024-01-25 20:10   ` Andy Caldwell [this message]
2024-01-25 21:25     ` [musl] RE: [EXTERNAL] " Rich Felker
2024-01-26 17:13       ` Andy Caldwell
2024-01-26 17:27         ` Rich Felker
2024-01-26 19:12           ` Andy Caldwell
2024-01-26 19:57             ` Rich Felker
2024-01-26 20:16               ` Andy Caldwell
2024-01-27 11:04                 ` Szabolcs Nagy
2024-01-27 12:58                   ` Alexander Monakov
2024-01-27 14:56                     ` Rich Felker
2024-01-27 19:20                       ` Szabolcs Nagy

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=AS4PR83MB05462D2FE519A65787A596DACB7A2@AS4PR83MB0546.EURPRD83.prod.outlook.com \
    --to=andycaldwell@microsoft.com \
    --cc=musl@lists.openwall.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.
Code repositories for project(s) associated with this public inbox

	https://git.vuxu.org/mirror/musl/

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).