[parted-devel] [PATCH 1/2] do not discard bootcode from extended partition

[Petr Uzel]

On Fri, Aug 21, 2009 at 05:09:00PM +0200, Jim Meyering wrote:
> Petr Uzel wrote:
> > * libparted/labels/dos.c (write_ext_table): Do not discard
> >   bootcode from extended partition on msdos label when some of
> >   the logical partitions are changed
> 
> Hi Petr,
> Thanks for working on this.

Hi Jim,
Thanks for the comments - I agree with all of them.
Since the patch has been already applied, the question is, whether
I should
a) revert the applied patch and post a new one addressing the issues
b) leave the applied patch and fix the issues in other patches

In the latter case I don't know how to fix the changelog.

So which way do you prefer?


> ...
> > diff --git a/libparted/labels/dos.c b/libparted/labels/dos.c
> > index 1d4c2dd..7b0c6d6 100644
> > --- a/libparted/labels/dos.c
> > +++ b/libparted/labels/dos.c
> > @@ -1027,7 +1027,8 @@ write_ext_table (const PedDisk* disk,
> >                   PedSector sector, const PedPartition* logical)
> >  {
> >  	PedPartition*		part;
> > -	PedSector		lba_offset;
> > +	PedSector			lba_offset;
> > +	void*				s;
> 
> Please do not modify unrelated lines like the lba_offset declaration.
> Also, please leave the declaration of "s" back just before the first use.
> That helps readability/maintainability, since then, there is
> no risk that someone will use it between the declaration and first use.
> 
> >
> >  	PED_ASSERT (disk != NULL, return 0);
> >  	PED_ASSERT (ped_disk_extended_partition (disk) != NULL, return 0);
> > @@ -1035,10 +1036,11 @@ write_ext_table (const PedDisk* disk,
> >
> >  	lba_offset = ped_disk_extended_partition (disk)->geom.start;
> >
> > -	void *s = ped_calloc (disk->dev->sector_size);
> > -	if (s == NULL)
> > +	if (!ptt_read_sector (disk->dev, sector, &s))
> >  		return 0;
> > +
> >  	DosRawTable *table = s;
> > +	memset(&(table->partitions), 0, 4 * sizeof(DosRawPartition));
> 
> Please don't hard-code constants like that.
> The 3rd argument above should be "sizeof (table->partitions)".
> 
> >  	table->magic = PED_CPU_TO_LE16 (MSDOS_MAGIC);
> >
> >  	int ok = 0;
> > @@ -1073,10 +1075,15 @@ static int
> 
> Your log doesn't mention this function.
> Yet it is no longer officially writing an empty table.
> Such a change in semantics deserves an explanatory comment
> in the code as well as mention in the log.
> 
> >  write_empty_table (const PedDisk* disk, PedSector sector)
> >  {
> >  	DosRawTable		table;
> > +	void*			table_sector;
> >
> >  	PED_ASSERT (disk != NULL, return 0);
> >
> > -	memset (&table, 0, sizeof (DosRawTable));
> > +	if (ptt_read_sector (disk->dev, sector, &table_sector)) {
> > +		memcpy (&table, table_sector, sizeof(DosRawTable));
> 
> Please do not use "sizeof(TYPE)" when you can instead use
> "sizeof variable".  The latter is more maintainable.
> Here, you should use "sizeof table".
> 
> > +		free(table_sector);
> > +	}
> > +	memset (&(table.partitions), 0, 4 * sizeof(DosRawPartition));
> 
> Don't hard-code 4, as above.

-- 
Best regards / s pozdravem

Petr Uzel, openSUSE Community Multiplier Team
-----------------------------------------------------------------
SUSE LINUX, s.r.o.                          e-mail: puzel at suse.cz
Lihovarská 1060/12                          http://www.suse.cz
190 00 Prague 9, CR                             
-------------- next part --------------
A non-text attachment was scrubbed...
Name: not available
Type: application/pgp-signature
Size: 198 bytes
Desc: not available
URL: <http://lists.alioth.debian.org/pipermail/parted-devel/attachments/20090825/487a08e0/attachment.pgp>