linux/drivers/ata, branch v4.9

libata-scsi: disable SCT Write Same for the moment

2016-12-07T21:29:09Z

SCT Write Same support had been introduced with commit 7b2030942859 ("libata: Add support for SCT Write Same") Some problems, namely excessive userspace segfaults, had been reported at http://lkml.kernel.org/r/20160908192736.GA4356@gmail.com This lead to commit 0ce1b18c42a5 ("libata: Some drives failing on SCT Write Same") which strived to disable SCT Write Same on !ZAC devices. Due to the way this was done and to the logic in sd_config_write_same(), this didn't work for those devices that have ->max_ws_blocks > SD_MAX_WS10_BLOCKS: for these, ->no_write_same and ->max_write_same_sectors would still be non-zero, but ->ws10 == ->ws16 == 0. This would cause sd_setup_write_same_cmnd() to demultiplex REQ_OP_WRITE_SAME requests to WRITE_SAME, and these in turn aren't supported by libata-scsi: EXT4-fs (dm-1): Delayed block allocation failed for inode 2625094 at logical offset 2032 with max blocks 2 with error 121 EXT4-fs (dm-1): This should not happen!! Data will be lost 121 == EREMOTEIO is what scsi_io_completion() asserts in case of invalid opcodes. Back to the original problem of userspace segfaults: this can be tracked down to ata_format_sct_write_same() overwriting the input page. Sometimes, this page is ZERO_PAGE(0) which ceases to be filled with zeros from that point on. Since ZERO_PAGE(0) is used for userspace .bss mappings, code of the following is doomed: static char *a = NULL; /* .bss */ ... if (a) *a = 'a'; This problem is not solved by disabling SCT Write Same for !ZAC devices only. It can certainly be fixed, but the final release is quite close -- so disable SCT Write Same for all ATA devices rather than introducing some SCT key buffer allocation schemes at this point. Fixes: 7b2030942859 ("libata: Add support for SCT Write Same") Signed-off-by: Nicolai Stange Signed-off-by: Tejun Heo

ata: sata_mv: check for errors when parsing nr-ports from dt

2016-11-29T16:35:06Z

If the nr-ports property is missing ata_host_alloc_pinfo is called with n_ports = 0. This results in host->ports[0] = NULL which later makes mv_init_host() oops when dereferencing this pointer. Instead be a bit more cooperative and fail the probing with an error message. Signed-off-by: Uwe Kleine-König Signed-off-by: Tejun Heo

ahci: always fall back to single-MSI mode

2016-11-21T16:06:57Z

Don't try to guess what the errors from pci_irq_alloc_vectors mean, as that's too fragile. Instead always try allocating a single vector when multi-MSI mode fails. This makes various intel Desktop and Laptop CPUs use MSI again. Signed-off-by: Christoph Hellwig Reported-by: Michael Marley Tested-by: Michael Marley Fixes: 0b9e2988ab22 ("ahci: use pci_alloc_irq_vectors") Signed-off-by: Tejun Heo

libata-scsi: Fixup ata_gen_passthru_sense()

2016-11-01T17:30:15Z

There's a typo in ata_gen_passthru_sense(), where the first byte would be overwritten incorrectly later on. Reported-by: Charles Machalow Signed-off-by: Hannes Reinecke Fixes: 11093cb1ef56 ("libata-scsi: generate correct ATA pass-through sense") Cc: stable@vger.kernel.org # v4.7+ Signed-off-by: Tejun Heo

ahci: fix the single MSI-X case in ahci_init_one

2016-10-25T15:43:07Z

We need to make sure hpriv->irq is set properly if we don't use per-port vectors, so switch from blindly assigning pdev->irq to using pci_irq_vector, which handles all interrupt types correctly. Signed-off-by: Christoph Hellwig Reported-by: Robert Richter Tested-by: Robert Richter Tested-by: David Daney Fixes: 0b9e2988ab22 ("ahci: use pci_alloc_irq_vectors") Signed-off-by: Tejun Heo

ahci: fix nvec check

2016-10-20T17:40:14Z

commit 17a51f12 ("ahci: only try to use multi-MSI mode if there is more than 1 port") lead to a case where nvec isn't initialized before it's used. Fix this by moving the check into the n_ports conditional. Reported-and-reviewed-by Colin Ian King Signed-off-by: Christoph Hellwig Signed-off-by: Tejun Heo

ahci: only try to use multi-MSI mode if there is more than 1 port

2016-10-19T17:18:16Z

We should only try to allocate multiple MSI or MSI-X vectors if the device actually has multiple ports. Otherwise pci_alloc_irq_vectors will return a single vector due to n_ports = 1, in which case we shouldn't set the AHCI_HFLAG_MULTI_MSI flag. Signed-off-by: Christoph Hellwig Fixes: 0b9e2988 ("ahci: use pci_alloc_irq_vectors") Reported-by: Emmanuel Benisty Tested-by: Emmanuel Benisty Signed-off-by: Tejun Heo

ahci: qoriq: Revert "ahci: qoriq: Disable NCQ on ls2080a SoC"

2016-09-30T08:28:51Z

This reverts commit 640847298e2b7f19 ("ahci: qoriq: Disable NCQ on ls2080a SoC") The erratum has been fixed in ls2080a v2.0 and later soc. In reality, customer will not get any ls2080a v1.0 soc. Neither apply to any products. So reverting this commit won't create any side effect. Blacklisting v2.0 could also be a option, but that needs to check the soc version which is not suitable in the driver. Signed-off-by: Tang Yuantian Signed-off-by: Tejun Heo

pata_at91: Use PTR_ERR_OR_ZERO rather than if(IS_ERR(...)) + PTR_ERR

2016-09-22T15:48:46Z

Signed-off-by: Harman Kalra Signed-off-by: Tejun Heo

ata: Replace BUG() with BUG_ON().

2016-09-22T15:46:52Z

Replace BUG() with BUG_ON(). Caught by coccinelle. Signed-off-by: Harman Kalra Signed-off-by: Tejun Heo