> Pondering about this, it's ATA_LBA according to the docs, specifying
> that the address is an LBA.
This is true for some commands, but not all. It gets used for other stuff
too.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to [EMAIL PROTECTED
On Fri, Mar 23, 2007 at 01:03:15PM -0700, Randy Dunlap wrote:
> > It's 0x40. Its a "command dependant bit" - no useful name.
>
> dependent. OK, thanks.
>
Hi,
Pondering about this, it's ATA_LBA according to the docs, specifying
that the address is an LBA.
Cheers,
Kyle
-
To unsubscribe
On Wed, Mar 28, 2007 at 10:54:31PM +0100, Alan Cox wrote:
> > Hm. I tried adding it in the eh code after ata_set_mode() in
> > ata_eh_recover(), which alters the problem slightly - hpa_sectors is
> > still 0, so the taskfile call is still failing, but now the system just
> > stops at around the
On Wed, Mar 28, 2007 at 10:54:31PM +0100, Alan Cox wrote:
> I wonder if the firmware is dying when we ask the disk to go zero sized
> rather than erroring politely. I'm not sure hth HPA sectors can come back
> as zero but we can be fairly sure 0 means "no HPA" in this case I guess ?
No, it seems
> Hm. I tried adding it in the eh code after ata_set_mode() in
> ata_eh_recover(), which alters the problem slightly - hpa_sectors is
> still 0, so the taskfile call is still failing, but now the system just
> stops at around the time that anything attempts to access sda with no
> errors other
On Wed, Mar 28, 2007 at 10:57:54AM +0100, Alan Cox wrote:
> Ok thanks. This is interesting as the only thing new it is doing is
> asking for the HPA size. Does I think explain the problem however: Can
> you move the HPA setting call to after the mode has been set and see if
> that makes the proble
On Wed, 28 Mar 2007 01:08:52 +0100
Matthew Garrett <[EMAIL PROTECTED]> wrote:
> On Fri, Mar 23, 2007 at 07:13:21PM +, Alan Cox wrote:
> > For reference this is what I am currently using with 2.6.21-rc4-mm1 and
> > it is working for all my test cases so far: Its basically Kyle's patch
> > with
On Wed, Mar 28, 2007 at 01:08:52AM +0100, Matthew Garrett wrote:
> ata3.01: ata_hpa_resize 1: sectors = 234441648, hpa_sectors = 0
^
Does this just indicate the lack of an hpa? If so, the
/* if no hpa, both should be equal */
comment
On Fri, Mar 23, 2007 at 07:13:21PM +, Alan Cox wrote:
> For reference this is what I am currently using with 2.6.21-rc4-mm1 and
> it is working for all my test cases so far: Its basically Kyle's patch
> with a libata switch to turn it on/off and some minor fixups from
> the original patch as po
Matthew Garrett wrote:
On Fri, Mar 23, 2007 at 07:13:21PM +, Alan Cox wrote:
+static int ata_ignore_hpa = 0;
+module_param_named(ignore_hpa, ata_ignore_hpa, int, 0644);
+MODULE_PARM_DESC(ignore_hpa, "Ignore HPA (0=off 1=on)");
I'm not sure I like the language here. "Ignore HPA" appears to
On Fri, Mar 23, 2007 at 07:13:21PM +, Alan Cox wrote:
> For reference this is what I am currently using with 2.6.21-rc4-mm1 and
> it is working for all my test cases so far: Its basically Kyle's patch
> with a libata switch to turn it on/off and some minor fixups from
> the original patch as po
Alan Cox wrote:
What is 0x40? can it be #defined (or enum-ed) instead of a magic
value? please? (more of same below)
>>> It's 0x40. Its a "command dependant bit" - no useful name.
>> dependent. OK, thanks.
>
> IDE is a bit like that. I'm amazed some of the command flags arent in
> la
On Fri, Mar 23, 2007 at 07:13:21PM +, Alan Cox wrote:
> +static int ata_ignore_hpa = 0;
> +module_param_named(ignore_hpa, ata_ignore_hpa, int, 0644);
> +MODULE_PARM_DESC(ignore_hpa, "Ignore HPA (0=off 1=on)");
I'm not sure I like the language here. "Ignore HPA" appears to mean
"Explicitly dis
On Fri, 23 Mar 2007 12:22:32 -0700 (PDT)
David Miller <[EMAIL PROTECTED]> wrote:
> From: Alan Cox <[EMAIL PROTECTED]>
> Date: Fri, 23 Mar 2007 20:08:19 +
>
> > u64 is always unsigned long long (and its debug anyway)
>
> It's plain "unsigned long" on sparc64 and some other 64-bit platforms.
> > > What is 0x40? can it be #defined (or enum-ed) instead of a magic
> > > value? please? (more of same below)
> >
> > It's 0x40. Its a "command dependant bit" - no useful name.
>
> dependent. OK, thanks.
IDE is a bit like that. I'm amazed some of the command flags arent in
latin.
> Alrea
On Fri, 23 Mar 2007 20:08:19 + Alan Cox wrote:
> > > +static int ata_ignore_hpa = 0;
> >
> > Don't init to 0. Not needed, bloats binary files.
>
> It'll be one for the final release 8)
>
> > > +module_param_named(ignore_hpa, ata_ignore_hpa, int, 0644);
> > > +MODULE_PARM_DESC(ignore_hpa, "
From: Alan Cox <[EMAIL PROTECTED]>
Date: Fri, 23 Mar 2007 20:08:19 +
> u64 is always unsigned long long (and its debug anyway)
It's plain "unsigned long" on sparc64 and some other 64-bit platforms.
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a mess
> > +static int ata_ignore_hpa = 0;
>
> Don't init to 0. Not needed, bloats binary files.
It'll be one for the final release 8)
> > +module_param_named(ignore_hpa, ata_ignore_hpa, int, 0644);
> > +MODULE_PARM_DESC(ignore_hpa, "Ignore HPA (0=off 1=on)");
>
> So 1 = on = ignore, right?
Yes.
>
On Fri, 23 Mar 2007 19:13:21 + Alan Cox wrote:
> For reference this is what I am currently using with 2.6.21-rc4-mm1 and
> it is working for all my test cases so far: Its basically Kyle's patch
> with a libata switch to turn it on/off and some minor fixups from
> the original patch as posted
>
For reference this is what I am currently using with 2.6.21-rc4-mm1 and
it is working for all my test cases so far: Its basically Kyle's patch
with a libata switch to turn it on/off and some minor fixups from
the original patch as posted
diff -u --new-file --recursive --exclude-from /usr/src/exclu
20 matches
Mail list logo
