Tejun Heo wrote:
Ryan Power wrote:
I'm also experiencing this problem on a with the sata_nv module. I
suspect that it may at least not be controller dependant, as it's also
occurring on my silicon image controller (sata_sil, SiI3512). It looks
like something is zeroing out part of the return registers for the
overall smart status.
Tested with Maxtor 6L300S0 and Western Digital WD3200JS drives on both
the sil and nv controllers. This was working in 2.6.21.6, but I haven't
had a chance to try and isolate the problem any further.
Hmmm... Incidentally, Jeff told me yesterday he was seeing the same
problem on sil. Maybe it's something in the core. I'll investigate.
What about attached patch? Seems to make my mcp61 happy... (it is sent
from Mozilla, so patch below is most probably whitespace challenged...
attachment should be fine)
Petr
Fix reported task file values in sense data
ata_tf_read was setting HOB bit when lba48 command was submitted, but
was not clearing it before reading "normal" data. Maybe it would be
better to just clear HOB bit immediately after reading upper halves
for lba48 command, but I just decided to clear HOB bit in each
ata_tf_read...
Signed-off-by: Petr Vandrovec <petr@xxxxxxxxxxxxxx>
---
commit de1eff411670a3e5dfadcc281754cb26035779fa
tree 5fbcc8f667a34aa9688ce3aa19b9c92944a12862
parent c2e8a9b937eb84d1b712874f458790d2df25641a
author Petr Vandrovec <petr@xxxxxxxxxxxxxx> Sat, 14 Jul 2007 05:11:03 -0700
committer root <root@xxxxxxxxxxxxxxxxxxxxxxx> Sat, 14 Jul 2007 05:11:03
-0700
drivers/ata/libata-sff.c | 1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index a74afea..507ab69 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -196,6 +196,7 @@ void ata_tf_read(struct ata_port *ap, struct
ata_taskfile *t
f)
{
struct ata_ioports *ioaddr = &ap->ioaddr;
+ iowrite8(tf->ctl, ioaddr->ctl_addr);
tf->command = ata_check_status(ap);
tf->feature = ioread8(ioaddr->error_addr);
tf->nsect = ioread8(ioaddr->nsect_addr);
Fix reported task file values in sense data
ata_tf_read was setting HOB bit when lba48 command was submitted, but
was not clearing it before reading "normal" data. Maybe it would be
better to just clear HOB bit immediately after reading upper halves
for lba48 command, but I just decided to clear HOB bit in each
ata_tf_read...
Signed-off-by: Petr Vandrovec <petr@xxxxxxxxxxxxxx>
---
commit de1eff411670a3e5dfadcc281754cb26035779fa
tree 5fbcc8f667a34aa9688ce3aa19b9c92944a12862
parent c2e8a9b937eb84d1b712874f458790d2df25641a
author Petr Vandrovec <petr@xxxxxxxxxxxxxx> Sat, 14 Jul 2007 05:11:03 -0700
committer root <root@xxxxxxxxxxxxxxxxxxxxxxx> Sat, 14 Jul 2007 05:11:03 -0700
drivers/ata/libata-sff.c | 1 +
1 files changed, 1 insertions(+), 0 deletions(-)
diff --git a/drivers/ata/libata-sff.c b/drivers/ata/libata-sff.c
index a74afea..507ab69 100644
--- a/drivers/ata/libata-sff.c
+++ b/drivers/ata/libata-sff.c
@@ -196,6 +196,7 @@ void ata_tf_read(struct ata_port *ap, struct ata_taskfile *tf)
{
struct ata_ioports *ioaddr = &ap->ioaddr;
+ iowrite8(tf->ctl, ioaddr->ctl_addr);
tf->command = ata_check_status(ap);
tf->feature = ioread8(ioaddr->error_addr);
tf->nsect = ioread8(ioaddr->nsect_addr);
