linux-kernel - Re: [PATCH 1/1] fat: improve sync performance by grouping writes revised again

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: <f46018bb0611011002h1b3b6e5fjdc6cc032a7503dbd@mail.gmail.com>
Date:	Wed, 1 Nov 2006 13:02:12 -0500
From:	"Holden Karau" <holden@...scanfly.ca>
To:	"Jörn Engel" <joern@...nheim.fh-wedel.de>
Cc:	"Josef Sipek" <jsipek@....cs.sunysb.edu>,
	hirofumi@...l.parknet.co.jp, linux-kernel@...r.kernel.org,
	"Holden Karau" <holdenk@...dros.com>,
	"akpm@...l.org" <akpm@...l.org>, linux-fsdevel@...r.kernel.org,
	"Nick Piggin" <nickpiggin@...oo.com.au>,
	"Matthew Wilcox" <matthew@....cx>
Subject: Re: [PATCH 1/1] fat: improve sync performance by grouping writes revised again

On 11/1/06, Jörn Engel <joern@...nheim.fh-wedel.de> wrote:
> On Wed, 1 November 2006 11:17:50 -0500, Holden Karau wrote:
> > +     c_bh = kmalloc(nr_bhs*(sbi->fats) , GFP_KERNEL);
> > +     if (NULL == c_bh) {
> > +             printk(KERN_CRIT "not enough memory to store pointers to FAT blocks, will not sync. Possible data loss\n");
> > +             err = -ENOMEM;
> > +             goto error;
> > +     }
>
> o I personally hate Yoda code ("Null the pointer is not, young Jedi").
> o Old code simply returned -ENOMEM without printk.  Assuming this was
>   sufficient before, the printk can be dropped.
Ok, I'll drop the printk
> o Some people prefer assigning err outside the condition.  It is
>   supposed to give slightly better code on i386, iirc.
>
> Result would be something like:
>         c_bh = kmalloc(...
>         err = -ENOMEM;
>         if (!c_bh)
>                 goto error;
That wouldn't work so well since we always return err, and possibly
slightly better code for i386 doesn't seem all that worth it.
>
> > +             for (n = 0 ; n < nr_bhs ;  n++ ) {
> > +                     c_bh[(copy-1)*nr_bhs+n] = sb_getblk(sb, backup_fat + bhs[n]->b_blocknr);
> > +                     /* If there is not enough memory, fall back to the old system */
> > +                     if (!c_bh[(copy-1)*nr_bhs+n]) {
> > +                             printk("fat: not enough memory for all blocks , syncing at %d\n" ,(copy-1)*nr_bhs+n);
>
> Whether this printk makes sense, I cannot tell.
I suppose I might as well drop it.
>
> > +                             fat_sync_bhs_optw( c_bh+i  , (copy-1)*nr_bhs+n-i-1 , wait );
> > +                             /* Free the now sync'd blocks */
> > +                             for (; i < (copy-1)*nr_bhs+n ; i++)
> > +                                     brelse(c_bh[i]);
> > +                             /* We try the same block again */
> > +                             c_bh[(copy-1)*nr_bhs+n] = sb_getblk(sb, backup_fat + bhs[n]->b_blocknr);
> > +                             if (!c_bh[(copy-1)*nr_bhs+n]) {
> > +                                     printk(KERN_CRIT "fat:not enough memory for block after existing blocks released. Possible data loss.\n");
Based on the same reasoning you provided, I should probably drop this one too.
> > +                                     err = -ENOMEM;
> > +                                     goto error;
> > +                             }
>
> As above.
I'll drop the printk, but the same holds true about err
>
> >  error:
> > +     if (NULL != c_bh) {
> > +             kfree(c_bh);
> > +     }
>
> kfree(NULL) works just fine.  You can remove the condition.
Thanks, I should have checked that :-)
>
> > +int fat_sync_bhs_optw(struct buffer_head **bhs, int nr_bhs ,int wait)
> >  {
> >       int i, err = 0;
> >
> >       ll_rw_block(SWRITE, nr_bhs, bhs);
> > -     for (i = 0; i < nr_bhs; i++) {
> > -             wait_on_buffer(bhs[i]);
> > -             if (buffer_eopnotsupp(bhs[i])) {
> > -                     clear_buffer_eopnotsupp(bhs[i]);
> > -                     err = -EOPNOTSUPP;
> > -             } else if (!err && !buffer_uptodate(bhs[i]))
> > -                     err = -EIO;
> > +     if (wait) {
> > +             for (i = 0; i < nr_bhs; i++) {
> > +                     wait_on_buffer(bhs[i]);
> > +                     if (buffer_eopnotsupp(bhs[i])) {
> > +                             clear_buffer_eopnotsupp(bhs[i]);
> > +                             err = -EOPNOTSUPP;
> > +                     } else if (!err && !buffer_uptodate(bhs[i]))
> > +                             err = -EIO;
> > +             }
> >       }
> > +
> >       return err;
> >  }
>
> You could keep the old indentation if your condition was changed to
>
>         if (!wait)
>                 return 0;
Sounds good.
>
> Jörn
>
> --
> You can take my soul, but not my lack of enthusiasm.
> -- Wally
>


-- 
Cell: 613-276-1645
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@...r.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/