linux-kernel - Re: [PATCH v2] bcachefs: Avoid a potential useless over memory allocation in bch2_prt

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: <pmyezdskyr3t2iyald4diibl6b7lttf46d6sskxfu2wq3pub7r@gmyezjb3wfeu>
Date: Sun, 18 Feb 2024 22:39:14 -0500
From: Kent Overstreet <kent.overstreet@...ux.dev>
To: Christophe JAILLET <christophe.jaillet@...adoo.fr>
Cc: Brian Foster <bfoster@...hat.com>, linux-kernel@...r.kernel.org, 
	kernel-janitors@...r.kernel.org, linux-bcachefs@...r.kernel.org
Subject: Re: [PATCH v2] bcachefs: Avoid a potential useless over memory
 allocation in bch2_prt_[v]printf()

On Sun, Feb 18, 2024 at 05:12:28PM +0100, Christophe JAILLET wrote:
> 2 issues related to incorrect available space in the output buffer
> computation may lead to some extra memory allocation.
> 
> 
> First: vsnprintf() takes the size of the buffer, *including* the space for
> the trailing null.
> 
> But printbuf_remaining() returns the number of characters we can print
> to the output buffer, *excluding* the terminating null.
> 
> So, use printbuf_remaining_size(), which includes the space for the
> terminating null.

nope, buggy.

passing printbuf_remaining_size() to snprintf() is correct, but
snprintf() returns the number of charecters wrote _excluding_ the null

> 
> 
> Second: if the return value of vsnprintf() is greater than or equal to the
> passed size, the resulting string is truncated.
> So, in order to see if some extra space is needed, the check needs to be
> changed.
> 
> Signed-off-by: Christophe JAILLET <christophe.jaillet@...adoo.fr>
> ---
> Un-tested
> 
> v2: - Use printbuf_remaining_size() instead of hand-writing it.  [Brian Foster]
>     - Reword the commit log, hoping it is clearer.
>     - Synch with -next-20240215
> 
> v1: https://lore.kernel.org/all/0f40108bed3e084057223bdbe32c4b37f8500ff3.1694845203.git.christophe.jaillet@wanadoo.fr/
> ---
>  fs/bcachefs/printbuf.c | 10 ++++++----
>  1 file changed, 6 insertions(+), 4 deletions(-)
> 
> diff --git a/fs/bcachefs/printbuf.c b/fs/bcachefs/printbuf.c
> index b27d22925929..8cee9c2f88c4 100644
> --- a/fs/bcachefs/printbuf.c
> +++ b/fs/bcachefs/printbuf.c
> @@ -55,9 +55,10 @@ void bch2_prt_vprintf(struct printbuf *out, const char *fmt, va_list args)
>  		va_list args2;
>  
>  		va_copy(args2, args);
> -		len = vsnprintf(out->buf + out->pos, printbuf_remaining(out), fmt, args2);
> +		len = vsnprintf(out->buf + out->pos, printbuf_remaining_size(out),
> +				fmt, args2);
>  		va_end(args2);
> -	} while (len + 1 >= printbuf_remaining(out) &&
> +	} while (len >= printbuf_remaining_size(out) &&
>  		 !bch2_printbuf_make_room(out, len + 1));
>  
>  	len = min_t(size_t, len,
> @@ -72,9 +73,10 @@ void bch2_prt_printf(struct printbuf *out, const char *fmt, ...)
>  
>  	do {
>  		va_start(args, fmt);
> -		len = vsnprintf(out->buf + out->pos, printbuf_remaining(out), fmt, args);
> +		len = vsnprintf(out->buf + out->pos, printbuf_remaining_size(out),
> +				fmt, args);
>  		va_end(args);
> -	} while (len + 1 >= printbuf_remaining(out) &&
> +	} while (len >= printbuf_remaining_size(out) &&
>  		 !bch2_printbuf_make_room(out, len + 1));
>  
>  	len = min_t(size_t, len,
> -- 
> 2.43.2
>