linux-kernel - Re: [PATCH v1] xarray: update document for error space returned by xarray normal API

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 for Android: free password hash cracker in your pocket

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

Message-ID: <20200720120310.GV12769@casper.infradead.org>
Date:   Mon, 20 Jul 2020 13:03:10 +0100
From:   Matthew Wilcox <willy@...radead.org>
To:     Wang Long <w@...qinren.net>
Cc:     linux-kernel@...r.kernel.org, linux-fsdevel@...r.kernel.org
Subject: Re: [PATCH v1] xarray: update document for error space returned by
 xarray normal API

On Mon, Jul 20, 2020 at 12:17:38PM +0800, Wang Long wrote:
> In the current xarray code, the negative value -1 and -4095 represented
> as an error.
> 
> xa_is_err(xa_mk_internal(-4095)) and xa_is_err(xa_mk_internal(-1))
> are all return true.
> 
> This patch update the document.

There are actually three distinct problems here, but none of them are
fixed by this patch.

The first is that there's no test-suite coverage for this.
The second is that xa_is_err() is checking against -MAX_ERRNO instead of
-1023.
The third is that the documentation isn't clear enough because the line
you're correcting is accurate.

I don't think any of these three problems is terribly urgent to fix.
The second is most important because it could lead to confusion between
an xa_node that happens to be allocated at the top of memory and an
error return, but I don't think there is ever a situation where we end
up checking a node entry for being an error entry.  I may be wrong.

The solution to the first problem probably looks like this:

+++ b/lib/test_xarray.c
@@ -81,6 +81,11 @@ static void *xa_store_order(struct xarray *xa, unsigned long index,
 
 static noinline void check_xa_err(struct xarray *xa)
 {
+       XA_BUG_ON(xa, xa_is_err(xa_mk_internal(0)));
+       XA_BUG_ON(xa, !xa_is_err(xa_mk_internal(-1)));
+       XA_BUG_ON(xa, !xa_is_err(xa_mk_internal(-1023)));
+       XA_BUG_ON(xa, xa_is_err(xa_mk_internal(-1024)));
+
        XA_BUG_ON(xa, xa_err(xa_store_index(xa, 0, GFP_NOWAIT)) != 0);
        XA_BUG_ON(xa, xa_err(xa_erase(xa, 0)) != 0);
 #ifndef __KERNEL__


> Signed-off-by: Wang Long <w@...qinren.net>
> ---
>  include/linux/xarray.h | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/include/linux/xarray.h b/include/linux/xarray.h
> index b4d70e7..0588fb9 100644
> --- a/include/linux/xarray.h
> +++ b/include/linux/xarray.h
> @@ -36,7 +36,7 @@
>   * 257: Zero entry
>   *
>   * Errors are also represented as internal entries, but use the negative
> - * space (-4094 to -2).  They're never stored in the slots array; only
> + * space (-4095 to -1).  They're never stored in the slots array; only
>   * returned by the normal API.
>   */
>  
> -- 
> 1.8.3.1
> 
> 
> 
>