linux-kernel - Re: [PATCH 3/5] xfs: automatic resource cleanup of for_each

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>] [day] [month] [year] [list]

Message-ID: <YRFEtK1CNr0Q+4nz@infradead.org>
Date:   Mon, 9 Aug 2021 16:07:32 +0100
From:   Christoph Hellwig <hch@...radead.org>
To:     "Darrick J. Wong" <djwong@...nel.org>
Cc:     linux-xfs@...r.kernel.org, linux-kernel@...r.kernel.org,
        ojeda@...nel.org, nathan@...nel.org, ndesaulniers@...gle.com,
        arnd@...db.de
Subject: Re: [PATCH 3/5] xfs: automatic resource cleanup of for_each_perag*

> +# Required for for_each_perag*
> +ccflags-y += -std=gnu99

I don't think it is up to an individual subsystem to pick a specific C
dialect.

I think the most important reason why the kernel sticks with gnu89 is
to avoid the misfeature of variable declarations in the middle of
blocks, and this change would lose it.

> +	xfs_agnumber_t		last_agno = 0;
>  	int			saved_error = 0;
>  	int			error = 0;
>  	LIST_HEAD		(buffer_list);
>  
>  	/* update secondary superblocks. */
> -	for_each_perag_from(mp, agno, pag) {
> +	for_each_perag_from(mp, iter, 1) {
>  		struct xfs_buf		*bp;
>  
> +		last_agno = iter.pag->pag_agno;

This is a really horrible API as it magically injects a local variable
in a macro.  It also leads to worse code generation and a small but
noticable increase in .text sie:

hch@...ck:~/work/xfs$ size xfs.o.*
   text	   data	    bss	    dec	    hex	filename
1521421	 301161	   1880	1824462	 1bd6ce	xfs.o.old
1521516	 301161	   1880	1824557	 1bd72d	xfs.o.new