From: "Jose R. Santos" <jrs@us.ibm.com>
Subject: Re: [PATCH] e2fsprogs: New bitmap and inode table allocation for
 FLEX_BG
Date: Fri, 11 Jan 2008 16:11:28 -0600
Message-ID: <20080111161128.23fac6bc@gara>
References: <20080111112806.1bbd9f38@gara>
	<20080111210104.GR3351@webber.adilger.int>
Mime-Version: 1.0
Content-Type: text/plain; charset=US-ASCII
Content-Transfer-Encoding: 7bit
Cc: Theodore Tso <tytso@mit.edu>,
	linux-ext4 <linux-ext4@vger.kernel.org>
To: Andreas Dilger <adilger@sun.com>
In-Reply-To: <20080111210104.GR3351@webber.adilger.int>
Sender: linux-ext4-owner@vger.kernel.org

On Fri, 11 Jan 2008 14:01:04 -0700
Andreas Dilger <adilger@sun.com> wrote:

> On Jan 11, 2008  11:28 -0600, Jose R. Santos wrote:
> > +blk_t ext2fs_flexbg_offset(ext2_filsys fs, dgrp_t group, int flexbg_size, 
> > +			   ext2fs_block_bitmap bmap, int offset, int size)

OK.
 
> Could you please add some comments for what this function is trying to do?
> 
> > +	last_grp = (group + (flexbg_size - (group % flexbg_size) - 1));
> 
> Is this the same as:
> 
> 	last_grp = group + (flexbg_size - 1) / flexbg_size * flexbg_size
> 	
> (i.e. trying to round up to the next even multiple of flexbg_size)?
> 
> Didn't we decide to have flexbg_size be a power-of-two value, so we could
> use shift and mask instead of divide and mod?  It's less of an issue because
> group is only a 32-bit value, I guess.

Yes, I fixes this in the kernel code but neglected to fix it on the here.
 
> > +	if (ext2fs_get_free_blocks(fs, start_blk, last_blk, 1, bmap, 
> > +				   &first_free))
> > +		return first_free;
> > +	
> > +	if (ext2fs_get_free_blocks(fs, first_free + offset, last_blk, size, 
> > +				   bmap, &first_free))
> > +		return first_free;
> > +
> > +	return first_free;
> > +}
> > +
> >  errcode_t ext2fs_allocate_group_table(ext2_filsys fs, dgrp_t group,
> >  				      ext2fs_block_bitmap bmap)
> >  {
> >  	errcode_t	retval;
> >  	blk_t		group_blk, start_blk, last_blk, new_blk, blk;
> > -	int		j;
> > +	dgrp_t		last_grp;
> > +	int		j, rem_grps, flexbg_size = 0;
> >  
> >  	group_blk = ext2fs_group_first_block(fs, group);
> >  	last_blk = ext2fs_group_last_block(fs, group);
> >  
> >  	if (!bmap)
> >  		bmap = fs->block_map;
> > +
> > +	if (EXT2_HAS_INCOMPAT_FEATURE (fs->super,
> > +				       EXT4_FEATURE_INCOMPAT_FLEX_BG)) {
> > +		flexbg_size = 1 << fs->super->s_log_groups_per_flex;
> > +		rem_grps = flexbg_size - (group % flexbg_size);
> 
> Hmm, no point in doing "groups % flexbg_size" if we have
> s_log_groups_per_flex.  Could do "groups & (flexbg_size - 1)" instead.
> 
> > @@ -101,7 +102,11 @@ int ext2fs_super_and_bgd_loc(ext2_filsys fs,
> > +	if (flex_bg_size) {
> > +		if ((group % flex_bg_size) == 0)
> > +			numblocks -= 2 + fs->inode_blocks_per_group;
> 
> Ditto.
> 
> > @@ -1045,6 +1046,20 @@ static void PRS(int argc, char *argv[])
> >  				exit(1);
> >  			}
> >  			break;
> > +		case 'G':
> > +			flex_bg_size = strtoul(optarg, &tmp, 0);
> > +			if (*tmp) {
> > +				com_err(program_name, 0,
> > +					_("Illegal number for Flex_BG size"));
> > +				exit(1);
> > +			}
> > +			if (flex_bg_size < 2 || 
> > +			    (flex_bg_size & (flex_bg_size-1)) != 0) {
> > +				com_err(program_name, 0,
> > +					_("Flex_BG size must be a power of 2"));
> > +				exit(1);
> > +			}
> > +			break;
> 
> We've been putting new options under "-E var=value"...  I don't know what
> Ted's thoughs are on using new option letters, though this one might qualify.

I thought this would qualify as a new option letter.  Waiting on input
from Ted.

> > @@ -1444,6 +1459,16 @@ static void PRS(int argc, char *argv[])
> >  		}
> >  	}
> >  
> > +	if(flex_bg_size) {
> 
> Space after "if ".

Will fix.
 
> > +		shift = 0;
> > +		tmp = flex_bg_size;
> > +		while ((tmp >>= 1UL) != 0UL)
> > +			shift++;
> 
> There isn't a "log2" function?

Couldn't find anything in lib or include.
 
> Cheers, Andreas
> --
> Andreas Dilger
> Sr. Staff Engineer, Lustre Group
> Sun Microsystems of Canada, Inc.
> 


-JRS