From a89611e8489ac24f371c9fd6fef6605b170b16ba Mon Sep 17 00:00:00 2001 From: Tejun Heo Date: Mon, 7 Apr 2008 22:47:19 +0900 Subject: [PATCH] ahci: use ata_wait_after_reset() instead of ata_sff_wait_ready() Implement ahci_check_ready() and replace ata_sff_wait_after_reset() with ata_wait_after_reset(). As ahci was faking TF access, this change doesn't result in any functional difference. Signed-off-by: Tejun Heo --- drivers/ata/ahci.c | 14 ++++++++++++-- 1 file changed, 12 insertions(+), 2 deletions(-) diff --git a/drivers/ata/ahci.c b/drivers/ata/ahci.c index 45a67a9ad8ab..7e251a2cbda5 100644 --- a/drivers/ata/ahci.c +++ b/drivers/ata/ahci.c @@ -1256,6 +1256,16 @@ static int ahci_exec_polled_cmd(struct ata_port *ap, int pmp, return 0; } +static int ahci_check_ready(struct ata_link *link) +{ + void __iomem *mmio = link->ap->ioaddr.cmd_addr; + u8 status = readl(mmio + PORT_TFDATA) & 0xFF; + + if (!(status & ATA_BUSY)) + return 1; + return 0; +} + static int ahci_do_softreset(struct ata_link *link, unsigned int *class, int pmp, unsigned long deadline) { @@ -1303,7 +1313,7 @@ static int ahci_do_softreset(struct ata_link *link, unsigned int *class, ahci_exec_polled_cmd(ap, pmp, &tf, 0, 0, 0); /* wait for link to become ready */ - rc = ata_sff_wait_after_reset(link, 1, deadline); + rc = ata_wait_after_reset(link, deadline, ahci_check_ready); /* link occupied, -ENODEV too is an error */ if (rc) { reason = "device not ready"; @@ -1426,7 +1436,7 @@ static int ahci_p5wdh_hardreset(struct ata_link *link, unsigned int *class, * have to be reset again. For most cases, this should * suffice while making probing snappish enough. */ - rc = ata_sff_wait_after_reset(link, 1, jiffies + 2 * HZ); + rc = ata_wait_after_reset(link, jiffies + 2 * HZ, ahci_check_ready); if (rc) ahci_kick_engine(ap, 0); -- 2.20.1