Re: [PATCH 5/5] libata: Add support for TRIM

[Mark Lord]

Mark Lord wrote:
> Matthew Wilcox wrote:
> > Signed-off-by: Matthew Wilcox <willy@xxxxxxxxxxxxxxx>
> > ---
> >  drivers/ata/libata-scsi.c |   46 
> > +++++++++++++++++++++++++++++++++++++++++++++
> >  1 files changed, 46 insertions(+), 0 deletions(-)
> >
> > diff --git a/drivers/ata/libata-scsi.c b/drivers/ata/libata-scsi.c
> > index b9747fa..d4c8b8b 100644
> > --- a/drivers/ata/libata-scsi.c
> > +++ b/drivers/ata/libata-scsi.c
> ...
>
> It works.  :)
>
> I'm using TRIM on a 120GB OCZ Vertex SSD here now,
..

I take that back now.  The command is apparently failing,
but we don't log the failure, and nor do we stop attempting
to continue to use it on future TRIMs.

I dug deeper into the kernel side, when my own attempts in hdparm
all failed, both using SG_IO directly and using the BLKDISCARD ioctl().

The commands are simply rejected by the drive, with status=0x51, err=0x04.

Everything *looks* okay on the Linux side, so I'm guessing that the OCZ
drive uses a vendor-unique opcode or protocol for it.  They do have it
in there for a windows tool, but no further info than that.

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