Tejun Heo wrote:
Back in 2.6.17-rc2, a libata module parameter was added for atapi_dmadir.
That's nice, but most SATA devices which need it will tell us about it
in their IDENTIFY PACKET response, as bit-15 of word-62 of the
returned data (as per ATA7, ATA8 specifications).
So for those
Tejun Heo wrote:
From: Albert Lee [EMAIL PROTECTED]
Treat zero xfer length as HSM violation. While at it, add
unlikely()'s to ATAPI ireason and transfer length checks.
tj: Formatted patch and added unlikely()'s.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
Signed-off-by: Tejun Heo
.
This patch fixes ATAPI regressions introduced by setting transfer
chunk size to allocation size.
Signed-off-by: Tejun Heo [EMAIL PROTECTED]
Cc: Albert Lee [EMAIL PROTECTED]
Acked-by: Albert Lee [EMAIL PROTECTED]
-
To unsubscribe from this list: send the line unsubscribe linux-ide
.
This patch fixes ATAPI regressions introduced by setting transfer
chunk size to allocation size.
Signed-off-by: Tejun Heo [EMAIL PROTECTED]
Cc: Albert Lee [EMAIL PROTECTED]
Sorry for the late reply. The new patch looks good.
However, I am a little worried about the following code segment
Tejun Heo wrote:
For misc ATAPI commands which transfer variable length data to the
host, overflow can occur due to application or hardware bug. Such
overflows can be ignored safely as long as overflow data is properly
drained. libata HSM implementation has this implemented in
Per Mark's comments, maybe all ATAPI tape drives need ATA_HORKAGE_STUCK_ERR.
This patch applys ATA_HORKAGE_STUCK_ERR for all ATAPI tape drives.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
Cc: Mark Lord [EMAIL PROTECTED]
---
diff -Nrup 01_ide_tape_stuck_err/drivers/ata/libata-core.c
the ERR bit and proceed sending the CDB.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
Cc: Alan Cox [EMAIL PROTECTED]
Cc: Mark Lord [EMAIL PROTECTED]
---
Revised per Alan, Mark and Tejun's comments. Tested ok with the Seagate
STT8000A tape drive.
Patch against the libata-dev tree.
diff -Nrup
Tejun Heo wrote:
Daniel Drake wrote:
Tejun Heo wrote:
4ata2.00: HSM violation: eh_analyze_tf: BUSY|DRQ
3ata2.00: exception Emask 0x0 SAct 0x0 SErr 0x0 action 0x2 frozen
3ata2.00: cmd a0/00:00:00:0a:00/00:00:00:00:00/a0 tag 0 cdb 0x5a data
10 in
4 res 58/00:02:00:0a:00/00:00:00:00:00/a0
Jeff Garzik wrote:
Albert Lee wrote:
Patch 2/2:
After reading the last pio data block, the HSM is waiting for device
to be idle, not waiting for the last interrupt.
This patch changes the state after PIO data-in to HSM_ST_IDLE instead
of HSM_ST_LAST for accuracy.
Signed-off-by: Albert
Mark Lord wrote:
I can read and write the tape successfully, and the written data matches
what is later read back from it.
Here's the hack I used. Not ready for mainline, but perhaps it will
help Alan or whomever come up with something nice for Jeff.
---
Sergei Shtylyov wrote:
Hello.
Mikael Pettersson wrote:
Previously I reported that the pata_pdc2027x PLL detection changes
in kernel 2.6.22 broke the driver on my PowerMac:
pata_pdc2027x: Invalid PLL input clock 1691742kHz, give up!
This is followed by a number of errors and speed
Minor cleanup to remove the unneeded rmb()s per Jeff's advice. Also removed the
pll_clock 0 check since pll_clock now guaranteed to be = 0 after Mikael's
patch.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
Tested ok on both x86 and ppc64, together with Mikael's patch.
diff -Nrup 01_mikael
pdc202xx_new.c driver...
Acked-by: Albert Lee [EMAIL PROTECTED]
--
albert
-
To unsubscribe from this list: send the line unsubscribe linux-ide in
the body of a message to [EMAIL PROTECTED]
More majordomo info at http://vger.kernel.org/majordomo-info.html
Jeff Garzik wrote:
Mikael Pettersson wrote:
Previously I reported that the pata_pdc2027x PLL detection changes
in kernel 2.6.22 broke the driver on my PowerMac:
pata_pdc2027x: Invalid PLL input clock 1691742kHz, give up!
This is followed by a number of errors and speed reduction
steps
Jeff Garzik wrote:
Albert Lee wrote:
The first rmb() is to make sure bccrl is read before bccrlv for later
(bccrl = bccrlv) check since both reading the same memory address.
That's already guaranteed without the rmb(), AFAICS.
Hmm, thanks for the advice. Will remove both rmb()s
Vasily Averin wrote:
Tejun Heo wrote:
[cc'ing Albert]
Vasily Averin wrote:
Tejun, Jeff
I've noticed that some scsi commands for DVD-drive attached to pata_via
successfully finishes without any delays but reports about TIMEOUT
condition. It
happens because of ATA_ERR bit is set in status
It seems irq_on() in ata_bus_reset() and ata_std_postreset()
are leftover of the EDD reset. Remove them.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 00_libata-dev/drivers/ata/libata-core.c
01_remove_leftover_irqon/drivers/ata/libata-core.c
--- 00_libata-dev/drivers/ata/libata
After checking the current implementations of freeze()/thaw(), it seems only
pdc_freeze()
does more than simple irq masking. Remove the DMA stop code from pdc_freeze().
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 01_remove_leftover_irqon/drivers/ata/sata_promise.c
.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
Considerations:
Polling PIO, the new user of freeze()/thaw(), might have higher calling
frequency to freeze()/thaw() than EH.
Can current implemention of freeze()/thaw() handle that?
We might need more testing on sata_nv, sata_promise, sata_sil
calls irq_on/irq_off explicitly. vsc_intr_mask_update() also removed
since now unused.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
The bitmasks used by vsc_intr_mask_update() are different to the bit masks
in vsc_irq_on() and vsc_irq_off(). I don't know which one is more correct.
Maybe we need
()
and such mechanism isn't much utilized.)
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 06_tf_load_cleanup/drivers/ata/libata-core.c
07_last_ctl_cleanup/drivers/ata/libata-core.c
--- 06_tf_load_cleanup/drivers/ata/libata-core.c2007-07-07
10:49:04.0 +0800
Tejun Heo wrote:
I like the whole series. Just a few nits.
* 9 and 10 need to be oen patch. As it currently stands, compile will
fail after 9.
* Please don't include Patch n/10 in message body.
* Add freeze/thaw to all, merge, kill freeze/thaw from all sequence is a
bit odd. It
Jeff Garzik wrote:
I have patches from Alan (pata_sis FIFO whack, pata_dma option), Tejun,
Albert and Kristen still to be reviewed. Will get to those on Friday,
after the July 4th US holiday.tions(-)
Just to be more specific, my to-review inbox contains:
Alan: pata_sis FIFO whack,
Sergei Shtylyov wrote:
Bartlomiej Zolnierkiewicz wrote:
typos fixed manually
Signed-off-by: Albert Lee [EMAIL PROTECTED]
Except for types in description:
Said he, while making his own typo. :-)
:)
--
albert
-
To unsubscribe from this list: send the line unsubscribe
Minor pio fixes:
1/2: move ata_altstatus() to pio data xfer functions
2/2: change the last state of pio read to HSM_ST_IDLE
(The previous remove unneeded ata_altstatus() from ata_hsm_qc_complete()
has been accepted.)
Patch against the libata-dev tree for your review.
-
To unsubscribe from this
Patch 1/2:
Move ata_altstatus() out from ata_hsm_move() to the pio data xfer functions
like ata_pio_sectors() and atapi_pio_bytes() where it makes more sense.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
atapi_send_cdb() already calls ata_altstatus() inside.
This patch makes ata_pio_sectors
For ATA, there are two levels of mechanism available to turn irq on/off.
- device level: nIEN bit in the control register. This masks INTRQ from the
device.
- host adapter level: some controller can mask out per-port irq from the host
adapter.
Currently various parts of libata deal with irq
Patch 2/10:
Currently there is -irq_on but no -irq_off.
Turning irq off is done via altering the nIEN bit of qc-tf, together with
tf_load().
This patch adds -irq_off for symmetry.
tf_load() and ata_qc_set_polling() will be fixed/removed in later patches.
Signed-off-by: Albert Lee [EMAIL
Patch 3/10:
Implement the newly added -irq_off in LLDDs.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 02_add_irq_off/drivers/ata/ahci.c
03_add_irq_off_lldd/drivers/ata/ahci.c
--- 02_add_irq_off/drivers/ata/ahci.c 2007-07-04 11:26:30.0 +0800
+++ 03_add_irq_off_lldd
now unused.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
Considerations:
HSM, the new user of freeze()/thaw(), will call freeze()/thaw() in higher
frequency than EH.
Can current implemention of freeze()/thaw() handle it?
diff -Nrup 04_convert_freeze/drivers/ata/libata-core.c
Patch 6/10:
Now that HSM polling code path uses freeze()/thaw() regardless of old EH or new
EH.
Add freeze()/thaw() to old EH LLDDs for the HSM polling code.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 05_convert_hsm_to_freeze/drivers/ata/pata_ixp4xx_cf.c
Patch 7/10:
After checking the current implementations of freeze()/thaw(), it seems only
pdc_freeze()
do more than simple irq masking. Remove the DMA stop code from pdc_freeze().
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 06_add_freeze_thaw_to_lldd/drivers/ata/sata_promise.c
off/on.
Remove the implicit nIEN handling from ata_tf_load().
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 07_sata_promise_freeze/drivers/ata/libata-sff.c
08_tfload_cleanup/drivers/ata/libata-sff.c
--- 07_sata_promise_freeze/drivers/ata/libata-sff.c 2007-07-04
13:12
Patch 9/10:
irq_on/irq_off are now only wrapped by freeze/thaw (and unused otherwise).
We can integrate freeze/thaw with irq_on/irq_off.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
This is for libata-core. The LLDDs will be fixed in the next patch.
diff -Nrup 08_tfload_cleanup/drivers/ata
Alan Cox wrote:
On Fri, Jun 29, 2007 at 05:34:36PM +1000, Andrew Hall wrote:
Further to this the PATA to SATA bridge being used in this case is:
http://www.jmicron.com/JM20330.html
..as you will see only PIO and UDMA modes are supported.
In which case their microcontroller in the middle
().
This patch calls gettimeofday() to mesure the time elapsed and
calculate the PLL input clock accordingly.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
Cc: Alan Cox [EMAIL PROTECTED]
---
(I have the Promise adapter at hand, so it's easier for me to verify.)
Attached please find the patch and the test result
In ata_hsm_qc_complete():
Calling ata_altstatus() after the qc is completed might race with next qc.
Remove it.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
Jeff,
(Sorry for re-submitting this patch so late.)
The unneeded reading of alt_status might cause trouble when another command
doing
Robert de Rooy wrote:
Albert Lee wrote:
Mark Lord wrote:
...
Mmm.. I don't know about the first failure there,
but after that it gets into the stuck DRQ state
which libata makes no attempt to handle at present.
It seems the pata_pcmcia driver is using IRQ driven PIO. Maybe
Alan Cox wrote:
Reloading the pata_pdc2027x module then the problem goes away.
I guess maybe somehow mdelay() is not as precise as before...
It's not the most accurate of things once power management and the like
get invovled. HT doesn't help it either. You should have get much more
().
This patch calls gettimeofday() to mesure the time elapsed and
calculate the PLL input clock accordingly.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
Cc: Alan Cox [EMAIL PROTECTED]
---
Did more test. For mdelay(100) the usec_elapsed is usually 99287.
However, sometimes the usec_elapsed is 118934, longer
Mark Lord wrote:
Robert de Rooy wrote:
I did another try with libata pcmcia support using 2.6.22-rc5 which
already includes the nodata polling fix, in combination with
disable-dev_init_param-and-setxfermode-for-CFA.patch and the
timing-debug.patch
...
Jun 22 13:19:44 localhost
Mikael Pettersson wrote:
I tried (again) to convert an Intel i815EP-chipset
machine from IDE to libata, but libata still fails
to initialise the CDROM on the machine.
With 2.6.22-rc5, IDE says:
Uniform Multi-Platform E-IDE driver Revision: 7.00alpha2
ide: Assuming 33MHz system bus speed
Sergei Shtylyov wrote:
Hello.
Bahadir Balban wrote:
I have a PCI Promise TX2 Ultra133 controller with a harddisk on an ARM
platform (which is a barebone system with no BIOS). This setup used to
work with the old linux-ide drivers on 2.6.19
You were very lucky then bacause actually
For ATA, there are two levels of mechanism available to turn irq on/off.
- device level: nIEN bit in the control register. nIEN masks INTRQ from the
device.
- host adapter level: some controller can mask out per-port irq from the host
adapter.
Currently various parts of libata deal with irq
Patch 1/9:
It looks the calling of irq_on() in ata_bus_reset() and ata_std_postreset()
are leftover of the earlier EDD reset. Remove them.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 00_libata-dev/drivers/ata/libata-core.c
01_remove_leftover_irqon/drivers/ata/libata-core.c
Patch 2/9:
Currently there is irq_on() but no irq_off().
Turning irq off is done via altering the nIEN bit of qc-tf, together with
tf_load().
This patch adds irq_off() for symmetry.
tf_load() and ata_qc_set_polling() will be fixed/removed in later patches.
Signed-off-by: Albert Lee [EMAIL
Patch 3/9:
Minor patch to add the newly added -irq_off() to LLDDs.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 02_add_irq_off/drivers/ata/ahci.c
03_add_irq_off_lldd/drivers/ata/ahci.c
--- 02_add_irq_off/drivers/ata/ahci.c 2007-06-01 12:08:21.0 +0800
Patch 4/9:
Minor patch to call irq_off() from bmdma_freeze() to avoid duplicated code.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 03_add_irq_off_lldd/drivers/ata/libata-sff.c
04_convert_freeze/drivers/ata/libata-sff.c
--- 03_add_irq_off_lldd/drivers/ata/libata-sff.c
than
using the device nIEN bit.
Considerations:
1. the semantic of freeze()/thaw() maybe more than irq off/on?
2. HSM, the new user of freeze()/thaw(), will call freeze()/thaw() more
frequently than EH.
Can current implemention of freeze()/thaw() handle it?
Signed-off-by: Albert Lee
-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 06_add_freeze_thaw_to_lldd/drivers/ata/sata_promise.c
07_sata_promise_freeze/drivers/ata/sata_promise.c
--- 06_add_freeze_thaw_to_lldd/drivers/ata/sata_promise.c 2007-06-11
17:23:47.0 +0800
+++ 07_sata_promise_freeze/drivers/ata
() are provided.
Since EH/HSM now call freeze()/thaw() for irq off/on explicitly.
The implicit nIEN handling is removed from ata_tf_load().
The nIEN snoop codes are also removed from sata_vsc.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 07_sata_promise_freeze/drivers/ata/libata
Patch 9/9:
It seems -irq_on and -irq_off are now only wrapped by freeze()/thaw() and
unused elsewhere.
This patch
- Remove -irq_on and -irq_off.
- Rename -freeze and -thaw to irq_on() and irq_off() to be specific.
Hopefully the LLDDs need to implement only one irq on/off by either host
1/6: update protocol numbers
2/6: support PIO multi commands
3/6: map UDMA protocols
4/6: always enforce correct DEV bit
5/6: support ATAPI devices
6/6: update cached device parameters
Patch against the libata-dev tree for your review, thanks.
-
To unsubscribe from this list: send the line
Patch 1/6:
Update the ATA passthru protocol numbers according to the new spec.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 00_libata-dev/drivers/ata/libata-scsi.c
01_protocol_update/drivers/ata/libata-scsi.c
--- 00_libata-dev/drivers/ata/libata-scsi.c 2007-06-01 12:08
Patch 2/6:
support the pass through of PIO multi commands.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 01_protocol_update/drivers/ata/libata-scsi.c
02_pio_multi/drivers/ata/libata-scsi.c
--- 01_protocol_update/drivers/ata/libata-scsi.c2007-06-07
11:38:50.0
Patch 3/6:
Map the ATA passthru UDMA protocols to ATA_PROT_DMA.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
Don't know why SAT distinguishs UDMA Data In and UDMA Data Out from DMA.
These UDMA protocols only matter for on-the-wire-protocol and are mostly
transparent
to the software. Anyway
Patch 4/6:
Always enforce correct DEV bit since we know which drive the command
is targeted. SAT demands to ignore the DEV bit, too.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 03_udma_supp/drivers/ata/libata-scsi.c
04_dev_bit/drivers/ata/libata-scsi.c
--- 03_udma_supp/drivers
Patch 5/6:
Support ATA passthru to ATAPI devices.
Revised based on Mark's patch and Jeff's comments.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
Cc: Mark Lord [EMAIL PROTECTED]
---
After reading Mark's patch, I guess we can safely by-pass the
cdb_len check: the ATA_12/ATA_16 are for the libata
Patch 6/6:
INIT_DEV_PARAMS and SET_MULTI_MODE change the device parameters cached by
libata.
Re-read IDENTIFY DEVICE info and update the cached device paramters when seeing
these commands.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 05_supp_atapi/drivers/ata/libata-scsi.c
Minor pio fixes:
1/3: remove unneeded ata_altstatus() from ata_hsm_qc_complete()
2/3: move the ata_altstatus() to pio data xfer functions
3/3: change the last state of pio read to HSM_ST_IDLE
Patch against the libata-dev tree for your review.
-
To unsubscribe from this list: send the line
Patch 1/3:
In ata_hsm_qc_complete():
Calling ata_altstatus() after the qc completed looks wrong.
Remove it.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
After checking, it is leftover of ata_pio_block( ) in 2.6.17-rc5 of the
following merge:
http://git.kernel.org/?p=linux/kernel/git/jgarzik
Patch 2/3:
Move ata_altstatus() out from ata_hsm_move() to the pio data xfer functions
such as ata_pio_sectors() and atapi_pio_bytes() where it makes more sense.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
atapi_send_cdb() already calls ata_altstatus() inside.
This patch makes
Patch 3/3:
After reading the last pio block, the HSM is waiting for device
to be idle, not waiting for the last interrupt.
Change the state after PIO data-in to HSM_ST_IDLE instead of HSM_ST_LAST for
accuracy.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 02_move_altstatus
For ATA/CFA devices, libata prints out the device model and firmware revision.
Do the same for ATAPI devices.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
Sometimes the error happens after identify but before the
SCSI messages. Knowing the device model and firmware revision
could also help
Alan Cox wrote:
As previously discussed, the possible issue with this patch is:
Some ATA/ATAPI devices might be unhappy if the STATUS register
is read during data transfer (not sure if this is true or not).
(Patch 5/8 doesn't have such issue.)
Some older intel eats your disk if you do that.
1/8: fix the ata_altstatus() in ata_hsm_qc_complete()
2/8: move ata_altstatus() from ata_hsm_move() to pio data xfer functions
3/8: change the state after PIO data-in to HSM_ST_IDLE instead of HSM_ST_LAST
4/8: move and reduce locking to the pio data xfer functions
5/8: ack possibly unsolicited irq
patch 1/8:
In ata_hsm_qc_complete():
Calling ata_altstatus() after the qc completed looks unsafe.
Move it to be before completing the qc.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
Don't know what the ata_altstatus() is doing here?
Anyway, move it to be before ata_qc_complete().
diff -Nrup
patch 2/8:
Move ata_altstatus() out from ata_hsm_move() to the pio data xfer functions,
like ata_pio_sectors() and atapi_pio_bytes() that know better if
ata_altstatus() is needed.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
atapi_send_cdb() already did ata_altstatus() in itself.
This patch
transfer and
clearing of the ATA_PFLAG_HSM_WQ flag.
- The transfer of head is made to be multiple of 8-bytes such that
-data_xfer() could possibly utilize 32-bit pio/mmio.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
Cc: Alan Cox [EMAIL PROTECTED]
Cc: Tejun Heo [EMAIL PROTECTED]
---
The variable name
patch 5/8:
- Move the ATA_TFLAG_POLLING check from ata_interrupt to ata_host_intr
- Ack unsolicited INTRQ if polling and before the HSM accessing the port.
(e.g. some device asserts INTRQ even if polling and nIEN = 1.)
Signed-off-by: Albert Lee [EMAIL PROTECTED]
Cc: Alan Cox [EMAIL PROTECTED
patch 6/8:
- Delegate irq driven pio to workqueue.
- HSM_ST_LAST is kept in the interrupt since this is not CPU intensive.
(Mostly for DMA and non-data.)
Signed-off-by: Albert Lee [EMAIL PROTECTED]
Cc: Alan Cox [EMAIL PROTECTED]
Cc: Tejun Heo [EMAIL PROTECTED]
---
sata_sil is not changed
7/8:
Submitting tasks from ata_host_intr() breaks ata_port_flush_task().
This patch adds code to disable irq pio delegation to ata_port_flush_task().
Irq pio delegation is re-enabled when next qc is issued.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
Cc: Tejun Heo [EMAIL PROTECTED]
---
Don't know
patch 8/8: ack more unsolicited irq
Signed-off-by: Albert Lee [EMAIL PROTECTED]
Cc: Alan Cox [EMAIL PROTECTED]
Cc: Mark Lord [EMAIL PROTECTED]
Cc: Tejun Heo [EMAIL PROTECTED]
---
To address Alan's concern about the unsolicited irq (hopefully),
this patch tries to ack irq that happens after HSM
1/7: set the state after PIO data-in to HSM_ST_IDLE instead of HSM_ST_LAST
2/7: fix the ata_altstatus() in ata_hsm_qc_complete()
3/7: move ata_altstatus() out from ata_hsm_move() to the pio data xfer functions
4/7: move polling idle irq check to ata_host_intr()
5/7: move and reduce locking to the
patch 2/7:
Calling ata_altstatus() after the qc completed looks incorrect.
Move it to before the qc is completed.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 01_last_idle/drivers/ata/libata-core.c
02_flush_fix/drivers/ata/libata-core.c
--- 01_last_idle/drivers/ata/libata-core.c
patch 3/7:
move ata_altstatus() out from ata_hsm_move() to the pio data xfer functions.
Functions like ata_pio_sectors() and atapi_pio_bytes() know better if the
flush is needed.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 02_flush_fix/drivers/ata/libata-core.c
03_smart_flush
patch 4/7:
move the polling idle irq check from ata_interrupt() to ata_host_intr(),
where it makes more sense.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 03_smart_flush/drivers/ata/libata-core.c
04_polling_check/drivers/ata/libata-core.c
--- 03_smart_flush/drivers/ata/libata
the ATA_PFLAG_HSM_WQ flag
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 04_polling_check/drivers/ata/libata-core.c
05_narrow_lock/drivers/ata/libata-core.c
--- 04_polling_check/drivers/ata/libata-core.c 2007-05-11 10:25:09.0
+0800
+++ 05_narrow_lock/drivers/ata/libata-core.c2007
patch 6/7:
Push part of the irq driven pio out to workqueue.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
It seems this creates new race condition with ata_exec_internal_sg() and EH?
diff -Nrup 05_narrow_lock/drivers/ata/libata-core.c
06_irq_wq/drivers/ata/libata-core.c
--- 05_narrow_lock
patch 7/7:
ack unexpected INTRQ when polling.
(Some device asserts INTRQ even if polling and nIEN = 1.
http://bugzilla.kernel.org/show_bug.cgi?id=8441)
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
diff -Nrup 06_irq_wq/drivers/ata/libata-core.c
07_ack_random_irq/drivers/ata/libata-core.c
Tejun Heo wrote:
diff -Nrup 03_smart_flush/drivers/ata/libata-core.c
04_polling_check/drivers/ata/libata-core.c
--- 03_smart_flush/drivers/ata/libata-core.c 2007-05-11 10:24:19.0
+0800
+++ 04_polling_check/drivers/ata/libata-core.c2007-05-11
10:25:09.0 +0800
@@
Tejun Heo wrote:
Albert Lee wrote:
-static void ata_pio_sector(struct ata_queued_cmd *qc, int last)
+static void ata_pio_sector(struct ata_queued_cmd *qc, int last, int lock)
I think the naming of @lock is a bit confusing here. @clr_hsm_wq or
@last_sector, maybe?
How about
Tejun Heo wrote:
I guess it's about time to listen what Jeff thinks.
Albert, if Jeff agrees with pushing HSM or data transfer to workqueue,
are you interested in doing that? The HSM is your baby after all. :-)
Moving HSM task to the workqueue looks good idea. The mouse cursor of my
X
Tejun Heo wrote:
[cc'ing Bartlomiej and Mark, hi]
Hello, Albert.
Albert Lee wrote:
Problem:
Kernel got irq 5: nobody cared when using
libata + polling IDENTIFY + Promise 20275 adapter + Benq DW1620 drive.
Detail message available in bug 8441
(http://bugzilla.kernel.org
Tejun Heo wrote:
Hello, Albert.
Albert Lee wrote:
Tejun Heo wrote:
Also, this is a problem for not only IDENTIFY but all polling commands.
Yes, other command might also assert INTRQ during polling.
However, for the specific BENQ DW1620 drive, only IDENTIFY_PACKET_DEVICE
has such behavior
DEVICE,
even if nIEN = 1.
Proposed fix:
disable_irq() during polling IDENTIFY to work around, the same as what IDE
subsystem does.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
Some controller like Intel ICH4 is immune from the problem, with the same kernel
and the same Benq DW1620 drive. So
Stephen Clark wrote:
Hi,
When trying to boot FC7 RC3 on my HP N5430 laptop I get the following:
...
Write protecting the kernel read-only data: 845k
input: PS/2 Generic Mouse as /class/input/input2
SCSI subsystem initialized
libata version 2.20 loaded.
ACPI: Unable to derive IRQ for
patch 1/4: Reorder HSM_ST_FIRST
patch 2/4: Clear tf before doing request sense
patch 3/4: Limit max sector to 128 for TORiSAN DVD drives
patch 4/4: Limit ATAPI DMA to R/W commands only for TORiSAN DVD drives
(patch 2/4 revised per Tejun's advice.)
(patch 3/4 revised per Vlad's new test result.)
patch 1/4:
Reorder HSM_ST_FIRST, such that the task state transition is easier decoded
with human eyes.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
Patch against libata-dev tree, for your review, thanks.
diff -Nrup 00_libata-dev.ori/include/linux/libata.h
01_hsm_st/include/linux/libata.h
patch 2/4:
Clear tf before doing request sense.
This fixes the AOpen 56X/AKH timeout problem.
(http://bugzilla.kernel.org/show_bug.cgi?id=8244)
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
Per Tejun's advice to use result_tf instead.
Patch against libata-dev tree, for your review, thanks
patch 3/4:
The TORiSAN drive locks up when max sector == 256.
Limit max sector to 128 for the TORiSAN DRD-N216 drives.
(http://bugzilla.kernel.org/show_bug.cgi?id=6710)
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
Revised to use max sector = 128 per Vlad's new test result.
Patch against
patch 1/4: Reorder HSM_ST_FIRST
patch 2/4: Clear tf before doing request sense
patch 3/4: Limit max sector to 240 for TORiSAN DVD drives
patch 4/4: Limit ATAPI DMA to R/W commands only for TORiSAN DVD drives
-
To unsubscribe from this list: send the line unsubscribe linux-ide in
the body of a
patch 1/4:
Reorder HSM_ST_FIRST, such that the task state transition is easier decoded
with human eyes.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
Patch against libata-dev tree, for your review, thanks.
diff -Nrup 00_libata-dev.ori/include/linux/libata.h
01_hsm_st/include/linux/libata.h
patch 2/4:
Clear tf before doing request sense.
This fixes the AOpen 56X/AKH timeout problem.
(http://bugzilla.kernel.org/show_bug.cgi?id=8244)
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
Patch against libata-dev tree, for your review, thanks.
diff -Nrup 01_hsm_st/drivers/ata/libata-eh.c
patch 3/4:
The TORiSAN drive locks up when max sector == 256.
Limit max sector to 240 for TORiSAN DRD-N216 drives.
(http://bugzilla.kernel.org/show_bug.cgi?id=6710)
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
Patch against libata-dev tree, for your review, thanks.
diff -Nrup
patch 4/4:
Limit ATAPI DMA to R/W commands only for TORiSAN DRD-N216 DVD-ROM drives
(http://bugzilla.kernel.org/show_bug.cgi?id=6710)
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
I guess maybe other commands like READ_DVD_STRUCTURE and READ_CD also
work under ATAPI DMA, but not confirmed
Tejun Heo wrote:
Albert Lee wrote:
diff -Nrup 01_hsm_st/drivers/ata/libata-eh.c
02_aopen_rs/drivers/ata/libata-eh.c
--- 01_hsm_st/drivers/ata/libata-eh.c2007-03-23 16:56:13.0
+0800
+++ 02_aopen_rs/drivers/ata/libata-eh.c2007-03-31
01:11:01.0 +0800
@@ -991,18
patch 1/4: Reorder HSM_ST_FIRST
patch 2/4: Clear tf before doing request sense
patch 3/4: Limit max sector to 240 for TORiSAN DVD drives
patch 4/4: Limit ATAPI DMA to R/W commands only for TORiSAN DVD drives
(patch 2/4 revised per Tejun's advice.)
-
To unsubscribe from this list: send the line
patch 1/4:
Reorder HSM_ST_FIRST, such that the task state transition is easier decoded
with human eyes.
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
Patch against libata-dev tree, for your review, thanks.
diff -Nrup 00_libata-dev.ori/include/linux/libata.h
01_hsm_st/include/linux/libata.h
patch 2/4:
Clear tf before doing request sense.
This fixes the AOpen 56X/AKH timeout problem.
(http://bugzilla.kernel.org/show_bug.cgi?id=8244)
Signed-off-by: Albert Lee [EMAIL PROTECTED]
---
Revised per Tejun's advice to use result_tf instead.
Patch against libata-dev tree, for your review
1 - 100 of 181 matches
Mail list logo
