linux-kernel - Re: [PATCH v1 3/8] md/raid5: Move common stripe count increment code into __find

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: <Yk/P0e6E7uVLkMnZ@infradead.org>
Date:   Thu, 7 Apr 2022 23:01:53 -0700
From:   Christoph Hellwig <hch@...radead.org>
To:     Logan Gunthorpe <logang@...tatee.com>
Cc:     linux-kernel@...r.kernel.org, linux-raid@...r.kernel.org,
        Song Liu <song@...nel.org>, Shaohua Li <shli@...nel.org>,
        Guoqing Jiang <guoqing.jiang@...ux.dev>,
        Stephen Bates <sbates@...thlin.com>,
        Martin Oliveira <Martin.Oliveira@...eticom.com>,
        David Sloan <David.Sloan@...eticom.com>
Subject: Re: [PATCH v1 3/8] md/raid5: Move common stripe count increment code
 into __find_stripe()

On Thu, Apr 07, 2022 at 10:45:06AM -0600, Logan Gunthorpe wrote:
>  static struct stripe_head *__find_stripe(struct r5conf *conf, sector_t sector,
> -					 short generation)
> +					 short generation, int hash)
>  {
> +	int inc_empty_inactive_list_flag;
>  	struct stripe_head *sh;
>  
>  	pr_debug("__find_stripe, sector %llu\n", (unsigned long long)sector);
>  	hlist_for_each_entry(sh, stripe_hash(conf, sector), hash)
>  		if (sh->sector == sector && sh->generation == generation)
> -			return sh;
> +			goto found;
>  	pr_debug("__stripe %llu not in cache\n", (unsigned long long)sector);
>  	return NULL;
> +
> +found:
> +	if (!atomic_inc_not_zero(&sh->count)) {

There is a way on list iterators outside the loop body waging right
now.  So maybe just leave __find_stripe as-is and add a new
find_get_stripe that wraps it.  And then just return early when the
atomic_inc_not_zero dos not succeed and save on one level of
indentation.