libata updates
PATCH 1/1 libata updates
[just sent upstream; patch snipped due to size]
Please pull from 'upstream-linus' branch ofgit://git.kernel.org/pub/scm/linux/kernel/git/jgarzik/libata-dev.git
to receive the following updates:
drivers/scsi/Makefile | 2 drivers/scsi/ahci.c | 197 +- drivers/scsi/ata_piix.c | 392 +++-- drivers/scsi/libata-bmdma.c | 703 ++++++++++ drivers/scsi/libata-core.c | 2984 +++++++++++++++++++++----------------------- drivers/scsi/libata-scsi.c | 240 ++- drivers/scsi/libata.h | 3 drivers/scsi/pdc_adma.c | 6 drivers/scsi/sata_mv.c | 279 +++- drivers/scsi/sata_nv.c | 2 drivers/scsi/sata_promise.c | 129 + drivers/scsi/sata_qstor.c | 10 drivers/scsi/sata_sil.c | 128 - drivers/scsi/sata_sil24.c | 102 - drivers/scsi/sata_sis.c | 2 drivers/scsi/sata_svw.c | 2 drivers/scsi/sata_sx4.c | 25 drivers/scsi/sata_uli.c | 2 drivers/scsi/sata_via.c | 2 drivers/scsi/sata_vsc.c | 2 drivers/scsi/scsi_error.c | 7 include/linux/ata.h | 22 include/linux/libata.h | 180 ++ include/scsi/scsi_eh.h | 3 24 files changed, 3312 insertions(+), 2112 deletions(-)
Albert Lee:
libata CHS: LBA28/LBA48 optimization (revise #6)
Daniel Drake:
sata_promise: Support FastTrak TX4300/TX4310
Jeff Garzik:
[libata ata_piix] Fix ICH6/7 map value interpretation [libata sata_mv] add 6042 support, fix 60xx/50xx EDMA configuration [libata scsi] build fix for ATA_FLAG_IN_EH change [libata] Move PCI IDE BMDMA-related code to new file libata-bmdma.c.
libata: turn on ATAPI by default
Luke Kosewski:
[libata sata_promise] add correct read/write of hotplug registers for SATAII devices
Randy Dunlap:
Various libata documentation updates.
Tejun Heo:
libata: separate out ata_sata_print_link_status
ahci: separate out ahci_stop/start_engine
ahci: separate out ahci_dev_classify
ata_piix: fix MAP VALUE interpretation for for ICH6/7
libata: fold __ata_qc_complete() into ata_qc_free()
libata: make the owner of a qc responsible for freeing it
libata: fix ata_qc_issue() error handling
ahci: fix err_mask setting in ahci_host_intr
libata: add detailed AC_ERR_* flags
libata: return AC_ERR_* from issue functions
SCSI: export scsi_eh_finish_cmd() and scsi_eh_flush_done_q()
libata: implement and apply ata_eh_qc_complete/retry()
libata: create pio/atapi task queueing wrappers
ahci: stop engine during hard reset
ahci: add constants for SRST
libata: export ata_busy_sleep
libata: modify ata_dev_try_classify
libata: new ->probe_reset operation
libata: implement ata_drive_probe_reset()
libata: implement standard reset component operations and ->probe_reset
libata: implement ATA_FLAG_IN_EH port flag
libata: EH / pio tasks synchronization
libata: fix ata_std_probe_reset() SATA detection
libata: separate out sata_phy_resume() from sata_std_hardreset()
libata: add probeinit component operation to ata_drive_probe_reset()
libata: implement ata_std_probeinit()
libata: add ATA_QCFLAG_EH_SCHEDULED
libata: implement ata_scsi_timed_out()
libata: use ata_scsi_timed_out()
libata: kill NULL qc handling from ->eng_timeout callbacks
ahci: separate out ahci_fill_cmd_slot()
libata: make new reset act identical to ->phy_reset register-wise
libata: kill SError clearing in sata_std_hardreset().
sata_sil: convert to new reset mechanism
sata_sil24: convert to new reset mechanism
sata_sil24: add hardreset
libata: inline ata_qc_complete()
ahci: make ahci_fill_cmd_slot() take *pp instead of *ap
ahci: convert to new reset mechanism
libata: convert assert(X)'s in libata core layer to WARN_ON(!X)'s
libata: convert assert(xxx)'s in low-level drivers to WARN_ON(!xxx)'s
libata: kill assert() macro
libata: allow ->probe_reset to return ATA_DEV_UNKNOWN
ata_piix: kill spurious assignment in piix_sata_probe()
libata: implement ata_dev_id_c_string()
libata: use ata_dev_id_c_string()
libata: separate out ata_id_n_sectors()
libata: separate out ata_id_major_version()
libata: make ata_dump_id() take @id instead of @dev
libata: don't do EDD handling if ->probe_reset is used
libata: make ata_dev_knobble() per-device
libata: move cdb_len for host to device
libata: add per-device max_sectors
libata: kill sht->max_sectors
libata: rename ata_dev_id_[c_]string()
libata: update ata_dev_init_params()
libata: fix comment regarding setting cable type
ata_piix: convert pata to new reset mechanism
ata_piix: convert sata to new reset mechanism
libata: separate out ata_dev_read_id()
libata: kill ata_dev_reread_id()
sata_sil24: add a new PCI ID for SiI 3124
libata: kill illegal kfree(id)
sata_sil: remove unneeded ATA_FLAG_SRST from 3512 port info
libata: seperate out ata_class_present()
ata_piix: finer-grained port_info
ata_piix: add a couple of flags
ata_piix: implement proper port map
ata_piix: reimplement piix_sata_probe()
libata: convert dev->id to pointer
libata: separate out ata_dev_configure()
libata: fold ata_dev_config() into ata_dev_configure()
libata: reorganize ata_bus_probe()
ata_piix: rename PIIX_FLAG_IGN_PRESENT to PIIX_FLAG_IGNORE_PCS
sata_sil: replace sil_3112_m15w board id with sil_3112
sata_sil: use kzalloc
sata_sil: replace register address constants with sil_port[] entry
sata_sil: cosmetic flag/constant changes
libata: re-initialize parameters before configuring
libata: add @print_info argument to ata_dev_configure()
libata: implement ata_dev_revalidate()
libata: revalidate after transfer mode configuration
sata_sil24: fix mwdma_mask setting
libata: implement port_task
libata: convert pio_task and packet_task to port_task
libata: kill unused pio_task and packet_task
libata: rename ATA_FLAG_FLUSH_PIO_TASK to ATA_FLAG_FLUSH_PORT_TASK
libata: improve xfer mask constants and update ata_mode_string()
libata: add xfer_mask handling functions
libata: use ata_id_xfermask() in ata_dev_configure()
libata: use xfer_mask helpers in ata_dev_set_mode()
libata: reimplement ata_set_mode() using xfer_mask helpers
libata: kill unused xfer_mode functions
libata: fix missing classes[] initialization in ata_bus_probe()
sata_sil24: exit early from softreset if SStatus reports no device
sata_sil24: lengthen softreset timeout
ahci: enable prefetching for PACKET commands
libata: fix class handling in ata_bus_probe()
libata: check Word 88 validity in ata_id_xfer_mask()
libata: use local *id instead of dev->id in ata_dev_configure()
libata: move IDENTIFY info printing from ata_dev_read_id() to ata_dev_configure()
Quoted 2 lines · Jeff Garzik
Jeff Garzik:
[libata] Move PCI IDE BMDMA-related code to new filelibata-bmdma.c.
This is a most confusing choice as 80% of the file has nothing to do with bus mastering DMA, and a large amount of it has nothing to do with PCI either. Also lots of DMA stuff is in the drivers so all the new bus mastering drivers don't use bmdma.c
The split makes sense, the choice of name is peculiar, if not completely broken 8)
Alan
Quoted 6 lines · Alan Cox
Quoted 2 lines · Jeff Garzik
Jeff Garzik:
[libata] Move PCI IDE BMDMA-related code to new filelibata-bmdma.c.
Quoted 7 lines
This is a most confusing choice as 80% of the file has nothing to do with bus mastering DMA, and a large amount of it has nothing to do with PCI either. Also lots of DMA stuff is in the drivers so all the new bus mastering drivers don't use
bmdma.cThe split makes sense, the choice of name is peculiar, if not completely broken 8)
Peculiar? Probably... :)
In my [no doubt warped] brain, I equate the SFF-8038 interface to "PCI IDE BMDMA", and from there, view most hardware as a subset of PCI IDE BMDMA. It might not do DMA, might not be PCI, might not do irqs, but most PATA hardware seems able to be driven by a "bmdma driver". Thus, the name :)
Jeff
Quoted 5 lines · Jeff Garzik
In my [no doubt warped] brain, I equate the SFF-8038 interface to "PCI IDE BMDMA", and from there, view most hardware as a subset of PCI IDE BMDMA. It might not do DMA, might not be PCI, might not do irqs, but most PATA hardware seems able to be driven by a "bmdma driver". Thus, the name :)
Most of that file is ST-506 type register interfaces, only a tiny bit is SFF PCI IDE with DMA.
Expect some patches soon. I've got my tree merged against the git tree now, just needs some testing.
Alan
Quoted 10 lines · Alan Cox
Quoted 1 line · Jeff Garzik
In my [no doubt warped] brain, I equate the SFF-8038 interface to "PCI IDE BMDMA", and from there, view most hardware as a subset of PCI IDE BMDMA. It might not do DMA, might not be PCI, might not do irqs, but most PATA hardware seems able to be driven by a "bmdma driver". Thus, the name :)
Most of that file is ST-506 type register interfaces, only a tiny bit is SFF PCI IDE with DMA.
Expect some patches soon. I've got my tree merged against the git tree now, just needs some testing.
To be specific, are the patches against libata-dev.git 'upstream' branch?
(most should apply to Linus-vanilla branch, so regardless of the definition of "the git tree" it should be OK)
Jeff
Quoted 2 lines · Jeff Garzik
Alan Cox wrote:
To be specific, are the patches against libata-dev.git 'upstream' branch?
Yes