LKML Archive on lore.kernel.org
help / color / mirror / Atom feed
* [git patches] libata ACPI support
@ 2007-02-17 22:12 Jeff Garzik
  0 siblings, 0 replies; 10+ messages in thread
From: Jeff Garzik @ 2007-02-17 22:12 UTC (permalink / raw)
  To: Andrew Morton, Linus Torvalds; +Cc: linux-ide, LKML

This has been living in libata-dev#ALL (and thus -mm) for quite a while
now.

For both PATA and SATA, this helps at suspend/resume time.

For SATA, ACPI support mostly consists of taskfiles (ATA commands) that
the BIOS wants us to send to the system drive.  Most notably, if you
have set a hard drive password in BIOS, you cannot access your data
without ACPI support.  Given that libata has survived this long without
ACPI support, this does not appear to impact many people.

For PATA, it makes available some DMA/PIO timing functions that
occasionally cannot be found anywhere else.  There is at least one case
of this where NVIDIA needs ACPI's help to reliably handle various
DMA/PIO modes (search in lkml/linux-ide archives for "pata_acpi").

NOTE: It is difficult to tell from the diff below, but the module
parameter (or kernel command line arg) is "libata.noacpi" not the more
generic "noacpi" that the diff might mistakenly lead you to believe.
search for the associated module_param() code for further details.

Should this cause problems, we can easily turn it off.  But as months of
-mm testing seem to indicate, this hasn't caused much of a ripple, aside
from solving a few lingering problems.

Please pull from 'acpi' branch of
master.kernel.org:/pub/scm/linux/kernel/git/jgarzik/libata-dev.git acpi

to receive the following updates:

 Documentation/kernel-parameters.txt |    5 +
 drivers/ata/Kconfig                 |   13 +
 drivers/ata/Makefile                |    2 +-
 drivers/ata/libata-acpi.c           |  698 +++++++++++++++++++++++++++++++++++
 drivers/ata/libata-core.c           |   14 +
 drivers/ata/libata.h                |   15 +
 include/linux/libata.h              |    5 +
 7 files changed, 751 insertions(+), 1 deletions(-)
 create mode 100644 drivers/ata/libata-acpi.c

Fiodor Suietov (1):
      libata: wrong sizeof for BUFFER

Kristen Carlson Accardi (3):
      libata: ACPI and _GTF support
      libata: ACPI _SDD support
      libata: change order of _SDD/_GTF execution (resend #3)

diff --git a/Documentation/kernel-parameters.txt b/Documentation/kernel-parameters.txt
index abd575c..5bc8970 100644
--- a/Documentation/kernel-parameters.txt
+++ b/Documentation/kernel-parameters.txt
@@ -48,6 +48,7 @@ parameter is applicable:
 	ISAPNP	ISA PnP code is enabled.
 	ISDN	Appropriate ISDN support is enabled.
 	JOY	Appropriate joystick support is enabled.
+	LIBATA  Libata driver is enabled
 	LP	Printer support is enabled.
 	LOOP	Loopback device support is enabled.
 	M68k	M68k architecture is enabled.
@@ -1038,6 +1039,10 @@ and is between 256 and 4096 characters. It is defined in the file
 			emulation library even if a 387 maths coprocessor
 			is present.
 
+	noacpi		[LIBATA] Disables use of ACPI in libata suspend/resume
+			when set.
+			Format: <int>
+
 	noaliencache	[MM, NUMA] Disables the allcoation of alien caches in
 			the slab allocator.  Saves per-node memory, but will
 			impact performance on real NUMA hardware.
diff --git a/drivers/ata/Kconfig b/drivers/ata/Kconfig
index 3747457..4af0a4b 100644
--- a/drivers/ata/Kconfig
+++ b/drivers/ata/Kconfig
@@ -161,6 +161,19 @@ config SATA_INTEL_COMBINED
 	depends on IDE=y && !BLK_DEV_IDE_SATA && (SATA_AHCI || ATA_PIIX)
 	default y
 
+config SATA_ACPI
+	bool
+	depends on ACPI && PCI
+	default y
+	help
+	  This option adds support for SATA-related ACPI objects.
+	  These ACPI objects add the ability to retrieve taskfiles
+	  from the ACPI BIOS and write them to the disk controller.
+	  These objects may be related to performance, security,
+	  power management, or other areas.
+	  You can disable this at kernel boot time by using the
+	  option libata.noacpi=1
+
 config PATA_ALI
 	tristate "ALi PATA support (Experimental)"
 	depends on PCI && EXPERIMENTAL
diff --git a/drivers/ata/Makefile b/drivers/ata/Makefile
index cd096f0..74298af 100644
--- a/drivers/ata/Makefile
+++ b/drivers/ata/Makefile
@@ -66,4 +66,4 @@ obj-$(CONFIG_ATA_GENERIC)	+= ata_generic.o
 obj-$(CONFIG_PATA_LEGACY)	+= pata_legacy.o
 
 libata-objs	:= libata-core.o libata-scsi.o libata-sff.o libata-eh.o
-
+libata-$(CONFIG_SATA_ACPI) += libata-acpi.o
diff --git a/drivers/ata/libata-acpi.c b/drivers/ata/libata-acpi.c
new file mode 100644
index 0000000..b4e8be5
--- /dev/null
+++ b/drivers/ata/libata-acpi.c
@@ -0,0 +1,698 @@
+/*
+ * libata-acpi.c
+ * Provides ACPI support for PATA/SATA.
+ *
+ * Copyright (C) 2006 Intel Corp.
+ * Copyright (C) 2006 Randy Dunlap
+ */
+
+#include <linux/ata.h>
+#include <linux/delay.h>
+#include <linux/device.h>
+#include <linux/errno.h>
+#include <linux/kernel.h>
+#include <linux/acpi.h>
+#include <linux/libata.h>
+#include <linux/pci.h>
+#include "libata.h"
+
+#include <acpi/acpi_bus.h>
+#include <acpi/acnames.h>
+#include <acpi/acnamesp.h>
+#include <acpi/acparser.h>
+#include <acpi/acexcep.h>
+#include <acpi/acmacros.h>
+#include <acpi/actypes.h>
+
+#define SATA_ROOT_PORT(x)	(((x) >> 16) & 0xffff)
+#define SATA_PORT_NUMBER(x)	((x) & 0xffff)	/* or NO_PORT_MULT */
+#define NO_PORT_MULT		0xffff
+#define SATA_ADR_RSVD		0xffffffff
+
+#define REGS_PER_GTF		7
+struct taskfile_array {
+	u8	tfa[REGS_PER_GTF];	/* regs. 0x1f1 - 0x1f7 */
+};
+
+
+/**
+ * sata_get_dev_handle - finds acpi_handle and PCI device.function
+ * @dev: device to locate
+ * @handle: returned acpi_handle for @dev
+ * @pcidevfn: return PCI device.func for @dev
+ *
+ * This function is somewhat SATA-specific.  Or at least the
+ * PATA & SATA versions of this function are different,
+ * so it's not entirely generic code.
+ *
+ * Returns 0 on success, <0 on error.
+ */
+static int sata_get_dev_handle(struct device *dev, acpi_handle *handle,
+					acpi_integer *pcidevfn)
+{
+	struct pci_dev	*pci_dev;
+	acpi_integer	addr;
+
+	pci_dev = to_pci_dev(dev);	/* NOTE: PCI-specific */
+	/* Please refer to the ACPI spec for the syntax of _ADR. */
+	addr = (PCI_SLOT(pci_dev->devfn) << 16) | PCI_FUNC(pci_dev->devfn);
+	*pcidevfn = addr;
+	*handle = acpi_get_child(DEVICE_ACPI_HANDLE(dev->parent), addr);
+	if (!*handle)
+		return -ENODEV;
+	return 0;
+}
+
+/**
+ * pata_get_dev_handle - finds acpi_handle and PCI device.function
+ * @dev: device to locate
+ * @handle: returned acpi_handle for @dev
+ * @pcidevfn: return PCI device.func for @dev
+ *
+ * The PATA and SATA versions of this function are different.
+ *
+ * Returns 0 on success, <0 on error.
+ */
+static int pata_get_dev_handle(struct device *dev, acpi_handle *handle,
+				acpi_integer *pcidevfn)
+{
+	unsigned int bus, devnum, func;
+	acpi_integer addr;
+	acpi_handle dev_handle, parent_handle;
+	struct acpi_buffer buffer = {.length = ACPI_ALLOCATE_BUFFER,
+					.pointer = NULL};
+	acpi_status status;
+	struct acpi_device_info	*dinfo = NULL;
+	int ret = -ENODEV;
+	struct pci_dev *pdev = to_pci_dev(dev);
+
+	bus = pdev->bus->number;
+	devnum = PCI_SLOT(pdev->devfn);
+	func = PCI_FUNC(pdev->devfn);
+
+	dev_handle = DEVICE_ACPI_HANDLE(dev);
+	parent_handle = DEVICE_ACPI_HANDLE(dev->parent);
+
+	status = acpi_get_object_info(parent_handle, &buffer);
+	if (ACPI_FAILURE(status))
+		goto err;
+
+	dinfo = buffer.pointer;
+	if (dinfo && (dinfo->valid & ACPI_VALID_ADR) &&
+	    dinfo->address == bus) {
+		/* ACPI spec for _ADR for PCI bus: */
+		addr = (acpi_integer)(devnum << 16 | func);
+		*pcidevfn = addr;
+		*handle = dev_handle;
+	} else {
+		goto err;
+	}
+
+	if (!*handle)
+		goto err;
+	ret = 0;
+err:
+	kfree(dinfo);
+	return ret;
+}
+
+struct walk_info {		/* can be trimmed some */
+	struct device	*dev;
+	struct acpi_device *adev;
+	acpi_handle	handle;
+	acpi_integer	pcidevfn;
+	unsigned int	drivenum;
+	acpi_handle	obj_handle;
+	struct ata_port *ataport;
+	struct ata_device *atadev;
+	u32		sata_adr;
+	int		status;
+	char		basepath[ACPI_PATHNAME_MAX];
+	int		basepath_len;
+};
+
+static acpi_status get_devices(acpi_handle handle,
+				u32 level, void *context, void **return_value)
+{
+	acpi_status		status;
+	struct walk_info	*winfo = context;
+	struct acpi_buffer	namebuf = {ACPI_ALLOCATE_BUFFER, NULL};
+	char			*pathname;
+	struct acpi_buffer	buffer;
+	struct acpi_device_info	*dinfo;
+
+	status = acpi_get_name(handle, ACPI_FULL_PATHNAME, &namebuf);
+	if (status)
+		goto ret;
+	pathname = namebuf.pointer;
+
+	buffer.length = ACPI_ALLOCATE_BUFFER;
+	buffer.pointer = NULL;
+	status = acpi_get_object_info(handle, &buffer);
+	if (ACPI_FAILURE(status))
+		goto out2;
+
+	dinfo = buffer.pointer;
+
+	/* find full device path name for pcidevfn */
+	if (dinfo && (dinfo->valid & ACPI_VALID_ADR) &&
+	    dinfo->address == winfo->pcidevfn) {
+		if (ata_msg_probe(winfo->ataport))
+			ata_dev_printk(winfo->atadev, KERN_DEBUG,
+				":%s: matches pcidevfn (0x%llx)\n",
+				pathname, winfo->pcidevfn);
+		strlcpy(winfo->basepath, pathname,
+			sizeof(winfo->basepath));
+		winfo->basepath_len = strlen(pathname);
+		goto out;
+	}
+
+	/* if basepath is not yet known, ignore this object */
+	if (!winfo->basepath_len)
+		goto out;
+
+	/* if this object is in scope of basepath, maybe use it */
+	if (strncmp(pathname, winfo->basepath,
+	    winfo->basepath_len) == 0) {
+		if (!(dinfo->valid & ACPI_VALID_ADR))
+			goto out;
+		if (ata_msg_probe(winfo->ataport))
+			ata_dev_printk(winfo->atadev, KERN_DEBUG,
+				"GOT ONE: (%s) root_port = 0x%llx,"
+				" port_num = 0x%llx\n", pathname,
+				SATA_ROOT_PORT(dinfo->address),
+				SATA_PORT_NUMBER(dinfo->address));
+		/* heuristics: */
+		if (SATA_PORT_NUMBER(dinfo->address) != NO_PORT_MULT)
+			if (ata_msg_probe(winfo->ataport))
+				ata_dev_printk(winfo->atadev,
+					KERN_DEBUG, "warning: don't"
+					" know how to handle SATA port"
+					" multiplier\n");
+		if (SATA_ROOT_PORT(dinfo->address) ==
+			winfo->ataport->port_no &&
+		    SATA_PORT_NUMBER(dinfo->address) == NO_PORT_MULT) {
+			if (ata_msg_probe(winfo->ataport))
+				ata_dev_printk(winfo->atadev,
+					KERN_DEBUG,
+					"THIS ^^^^^ is the requested"
+					" SATA drive (handle = 0x%p)\n",
+					handle);
+			winfo->sata_adr = dinfo->address;
+			winfo->obj_handle = handle;
+		}
+	}
+out:
+	kfree(dinfo);
+out2:
+	kfree(pathname);
+
+ret:
+	return status;
+}
+
+/* Get the SATA drive _ADR object. */
+static int get_sata_adr(struct device *dev, acpi_handle handle,
+			acpi_integer pcidevfn, unsigned int drive,
+			struct ata_port *ap,
+			struct ata_device *atadev, u32 *dev_adr)
+{
+	acpi_status	status;
+	struct walk_info *winfo;
+	int		err = -ENOMEM;
+
+	winfo = kzalloc(sizeof(struct walk_info), GFP_KERNEL);
+	if (!winfo)
+		goto out;
+
+	winfo->dev = dev;
+	winfo->atadev = atadev;
+	winfo->ataport = ap;
+	if (acpi_bus_get_device(handle, &winfo->adev) < 0)
+		if (ata_msg_probe(ap))
+			ata_dev_printk(winfo->atadev, KERN_DEBUG,
+				"acpi_bus_get_device failed\n");
+	winfo->handle = handle;
+	winfo->pcidevfn = pcidevfn;
+	winfo->drivenum = drive;
+
+	status = acpi_get_devices(NULL, get_devices, winfo, NULL);
+	if (ACPI_FAILURE(status)) {
+		if (ata_msg_probe(ap))
+			ata_dev_printk(winfo->atadev, KERN_DEBUG,
+				"%s: acpi_get_devices failed\n",
+				__FUNCTION__);
+		err = -ENODEV;
+	} else {
+		*dev_adr = winfo->sata_adr;
+		atadev->obj_handle = winfo->obj_handle;
+		err = 0;
+	}
+	kfree(winfo);
+out:
+	return err;
+}
+
+/**
+ * do_drive_get_GTF - get the drive bootup default taskfile settings
+ * @ap: the ata_port for the drive
+ * @ix: target ata_device (drive) index
+ * @gtf_length: number of bytes of _GTF data returned at @gtf_address
+ * @gtf_address: buffer containing _GTF taskfile arrays
+ *
+ * This applies to both PATA and SATA drives.
+ *
+ * The _GTF method has no input parameters.
+ * It returns a variable number of register set values (registers
+ * hex 1F1..1F7, taskfiles).
+ * The <variable number> is not known in advance, so have ACPI-CA
+ * allocate the buffer as needed and return it, then free it later.
+ *
+ * The returned @gtf_length and @gtf_address are only valid if the
+ * function return value is 0.
+ */
+static int do_drive_get_GTF(struct ata_port *ap, int ix,
+			unsigned int *gtf_length, unsigned long *gtf_address,
+			unsigned long *obj_loc)
+{
+	acpi_status			status;
+	acpi_handle			dev_handle = NULL;
+	acpi_handle			chan_handle, drive_handle;
+	acpi_integer			pcidevfn = 0;
+	u32				dev_adr;
+	struct acpi_buffer		output;
+	union acpi_object 		*out_obj;
+	struct device			*dev = ap->host->dev;
+	struct ata_device		*atadev = &ap->device[ix];
+	int				err = -ENODEV;
+
+	*gtf_length = 0;
+	*gtf_address = 0UL;
+	*obj_loc = 0UL;
+
+	if (noacpi)
+		return 0;
+
+	if (ata_msg_probe(ap))
+		ata_dev_printk(atadev, KERN_DEBUG,
+			"%s: ENTER: ap->id: %d, port#: %d\n",
+			__FUNCTION__, ap->id, ap->port_no);
+
+	if (!ata_dev_enabled(atadev) || (ap->flags & ATA_FLAG_DISABLED)) {
+		if (ata_msg_probe(ap))
+			ata_dev_printk(atadev, KERN_DEBUG, "%s: ERR: "
+				"ata_dev_present: %d, PORT_DISABLED: %lu\n",
+				__FUNCTION__, ata_dev_enabled(atadev),
+				ap->flags & ATA_FLAG_DISABLED);
+		goto out;
+	}
+
+	/* Don't continue if device has no _ADR method.
+	 * _GTF is intended for known motherboard devices. */
+	if (!(ap->cbl == ATA_CBL_SATA)) {
+		err = pata_get_dev_handle(dev, &dev_handle, &pcidevfn);
+		if (err < 0) {
+			if (ata_msg_probe(ap))
+				ata_dev_printk(atadev, KERN_DEBUG,
+					"%s: pata_get_dev_handle failed (%d)\n",
+					__FUNCTION__, err);
+			goto out;
+		}
+	} else {
+		err = sata_get_dev_handle(dev, &dev_handle, &pcidevfn);
+		if (err < 0) {
+			if (ata_msg_probe(ap))
+				ata_dev_printk(atadev, KERN_DEBUG,
+					"%s: sata_get_dev_handle failed (%d\n",
+					__FUNCTION__, err);
+			goto out;
+		}
+	}
+
+	/* Get this drive's _ADR info. if not already known. */
+	if (!atadev->obj_handle) {
+		if (!(ap->cbl == ATA_CBL_SATA)) {
+			/* get child objects of dev_handle == channel objects,
+	 		 * + _their_ children == drive objects */
+			/* channel is ap->port_no */
+			chan_handle = acpi_get_child(dev_handle,
+						ap->port_no);
+			if (ata_msg_probe(ap))
+				ata_dev_printk(atadev, KERN_DEBUG,
+					"%s: chan adr=%d: chan_handle=0x%p\n",
+					__FUNCTION__, ap->port_no,
+					chan_handle);
+			if (!chan_handle) {
+				err = -ENODEV;
+				goto out;
+			}
+			/* TBD: could also check ACPI object VALID bits */
+			drive_handle = acpi_get_child(chan_handle, ix);
+			if (!drive_handle) {
+				err = -ENODEV;
+				goto out;
+			}
+			dev_adr = ix;
+			atadev->obj_handle = drive_handle;
+		} else {	/* for SATA mode */
+			dev_adr = SATA_ADR_RSVD;
+			err = get_sata_adr(dev, dev_handle, pcidevfn, 0,
+					ap, atadev, &dev_adr);
+		}
+		if (err < 0 || dev_adr == SATA_ADR_RSVD ||
+		    !atadev->obj_handle) {
+			if (ata_msg_probe(ap))
+				ata_dev_printk(atadev, KERN_DEBUG,
+					"%s: get_sata/pata_adr failed: "
+					"err=%d, dev_adr=%u, obj_handle=0x%p\n",
+					__FUNCTION__, err, dev_adr,
+					atadev->obj_handle);
+			goto out;
+		}
+	}
+
+	/* Setting up output buffer */
+	output.length = ACPI_ALLOCATE_BUFFER;
+	output.pointer = NULL;	/* ACPI-CA sets this; save/free it later */
+
+	/* _GTF has no input parameters */
+	err = -EIO;
+	status = acpi_evaluate_object(atadev->obj_handle, "_GTF",
+					NULL, &output);
+	if (ACPI_FAILURE(status)) {
+		if (ata_msg_probe(ap))
+			ata_dev_printk(atadev, KERN_DEBUG,
+				"%s: Run _GTF error: status = 0x%x\n",
+				__FUNCTION__, status);
+		goto out;
+	}
+
+	if (!output.length || !output.pointer) {
+		if (ata_msg_probe(ap))
+			ata_dev_printk(atadev, KERN_DEBUG, "%s: Run _GTF: "
+				"length or ptr is NULL (0x%llx, 0x%p)\n",
+				__FUNCTION__,
+				(unsigned long long)output.length,
+				output.pointer);
+		kfree(output.pointer);
+		goto out;
+	}
+
+	out_obj = output.pointer;
+	if (out_obj->type != ACPI_TYPE_BUFFER) {
+		kfree(output.pointer);
+		if (ata_msg_probe(ap))
+			ata_dev_printk(atadev, KERN_DEBUG, "%s: Run _GTF: "
+				"error: expected object type of "
+				" ACPI_TYPE_BUFFER, got 0x%x\n",
+				__FUNCTION__, out_obj->type);
+		err = -ENOENT;
+		goto out;
+	}
+
+	if (!out_obj->buffer.length || !out_obj->buffer.pointer ||
+	    out_obj->buffer.length % REGS_PER_GTF) {
+		if (ata_msg_drv(ap))
+			ata_dev_printk(atadev, KERN_ERR,
+				"%s: unexpected GTF length (%d) or addr (0x%p)\n",
+				__FUNCTION__, out_obj->buffer.length,
+				out_obj->buffer.pointer);
+		err = -ENOENT;
+		goto out;
+	}
+
+	*gtf_length = out_obj->buffer.length;
+	*gtf_address = (unsigned long)out_obj->buffer.pointer;
+	*obj_loc = (unsigned long)out_obj;
+	if (ata_msg_probe(ap))
+		ata_dev_printk(atadev, KERN_DEBUG, "%s: returning "
+			"gtf_length=%d, gtf_address=0x%lx, obj_loc=0x%lx\n",
+			__FUNCTION__, *gtf_length, *gtf_address, *obj_loc);
+	err = 0;
+out:
+	return err;
+}
+
+/**
+ * taskfile_load_raw - send taskfile registers to host controller
+ * @ap: Port to which output is sent
+ * @gtf: raw ATA taskfile register set (0x1f1 - 0x1f7)
+ *
+ * Outputs ATA taskfile to standard ATA host controller using MMIO
+ * or PIO as indicated by the ATA_FLAG_MMIO flag.
+ * Writes the control, feature, nsect, lbal, lbam, and lbah registers.
+ * Optionally (ATA_TFLAG_LBA48) writes hob_feature, hob_nsect,
+ * hob_lbal, hob_lbam, and hob_lbah.
+ *
+ * This function waits for idle (!BUSY and !DRQ) after writing
+ * registers.  If the control register has a new value, this
+ * function also waits for idle after writing control and before
+ * writing the remaining registers.
+ *
+ * LOCKING: TBD:
+ * Inherited from caller.
+ */
+static void taskfile_load_raw(struct ata_port *ap,
+				struct ata_device *atadev,
+				const struct taskfile_array *gtf)
+{
+	if (ata_msg_probe(ap))
+		ata_dev_printk(atadev, KERN_DEBUG, "%s: (0x1f1-1f7): hex: "
+			"%02x %02x %02x %02x %02x %02x %02x\n",
+			__FUNCTION__,
+			gtf->tfa[0], gtf->tfa[1], gtf->tfa[2],
+			gtf->tfa[3], gtf->tfa[4], gtf->tfa[5], gtf->tfa[6]);
+
+	if ((gtf->tfa[0] == 0) && (gtf->tfa[1] == 0) && (gtf->tfa[2] == 0)
+	    && (gtf->tfa[3] == 0) && (gtf->tfa[4] == 0) && (gtf->tfa[5] == 0)
+	    && (gtf->tfa[6] == 0))
+		return;
+
+	if (ap->ops->qc_issue) {
+		struct ata_taskfile tf;
+		unsigned int err;
+
+		ata_tf_init(atadev, &tf);
+
+		/* convert gtf to tf */
+		tf.flags |= ATA_TFLAG_ISADDR | ATA_TFLAG_DEVICE; /* TBD */
+		tf.protocol = atadev->class == ATA_DEV_ATAPI ?
+			ATA_PROT_ATAPI_NODATA : ATA_PROT_NODATA;
+		tf.feature = gtf->tfa[0];	/* 0x1f1 */
+		tf.nsect   = gtf->tfa[1];	/* 0x1f2 */
+		tf.lbal    = gtf->tfa[2];	/* 0x1f3 */
+		tf.lbam    = gtf->tfa[3];	/* 0x1f4 */
+		tf.lbah    = gtf->tfa[4];	/* 0x1f5 */
+		tf.device  = gtf->tfa[5];	/* 0x1f6 */
+		tf.command = gtf->tfa[6];	/* 0x1f7 */
+
+		err = ata_exec_internal(atadev, &tf, NULL, DMA_NONE, NULL, 0);
+		if (err && ata_msg_probe(ap))
+			ata_dev_printk(atadev, KERN_ERR,
+				"%s: ata_exec_internal failed: %u\n",
+				__FUNCTION__, err);
+	} else
+		if (ata_msg_warn(ap))
+			ata_dev_printk(atadev, KERN_WARNING,
+				"%s: SATA driver is missing qc_issue function"
+				" entry points\n",
+				__FUNCTION__);
+}
+
+/**
+ * do_drive_set_taskfiles - write the drive taskfile settings from _GTF
+ * @ap: the ata_port for the drive
+ * @atadev: target ata_device
+ * @gtf_length: total number of bytes of _GTF taskfiles
+ * @gtf_address: location of _GTF taskfile arrays
+ *
+ * This applies to both PATA and SATA drives.
+ *
+ * Write {gtf_address, length gtf_length} in groups of
+ * REGS_PER_GTF bytes.
+ */
+static int do_drive_set_taskfiles(struct ata_port *ap,
+		struct ata_device *atadev, unsigned int gtf_length,
+		unsigned long gtf_address)
+{
+	int			err = -ENODEV;
+	int			gtf_count = gtf_length / REGS_PER_GTF;
+	int			ix;
+	struct taskfile_array	*gtf;
+
+	if (ata_msg_probe(ap))
+		ata_dev_printk(atadev, KERN_DEBUG,
+			"%s: ENTER: ap->id: %d, port#: %d\n",
+			__FUNCTION__, ap->id, ap->port_no);
+
+	if (noacpi || !(ap->cbl == ATA_CBL_SATA))
+		return 0;
+
+	if (!ata_dev_enabled(atadev) || (ap->flags & ATA_FLAG_DISABLED))
+		goto out;
+	if (!gtf_count)		/* shouldn't be here */
+		goto out;
+
+	if (gtf_length % REGS_PER_GTF) {
+		if (ata_msg_drv(ap))
+			ata_dev_printk(atadev, KERN_ERR,
+				"%s: unexpected GTF length (%d)\n",
+				__FUNCTION__, gtf_length);
+		goto out;
+	}
+
+	for (ix = 0; ix < gtf_count; ix++) {
+		gtf = (struct taskfile_array *)
+			(gtf_address + ix * REGS_PER_GTF);
+
+		/* send all TaskFile registers (0x1f1-0x1f7) *in*that*order* */
+		taskfile_load_raw(ap, atadev, gtf);
+	}
+
+	err = 0;
+out:
+	return err;
+}
+
+/**
+ * ata_acpi_exec_tfs - get then write drive taskfile settings
+ * @ap: the ata_port for the drive
+ *
+ * This applies to both PATA and SATA drives.
+ */
+int ata_acpi_exec_tfs(struct ata_port *ap)
+{
+	int		ix;
+	int		ret =0;
+	unsigned int	gtf_length;
+	unsigned long	gtf_address;
+	unsigned long	obj_loc;
+
+	if (noacpi)
+		return 0;
+
+	for (ix = 0; ix < ATA_MAX_DEVICES; ix++) {
+		if (!ata_dev_enabled(&ap->device[ix]))
+			continue;
+
+		ret = do_drive_get_GTF(ap, ix,
+				&gtf_length, &gtf_address, &obj_loc);
+		if (ret < 0) {
+			if (ata_msg_probe(ap))
+				ata_port_printk(ap, KERN_DEBUG,
+					"%s: get_GTF error (%d)\n",
+					__FUNCTION__, ret);
+			break;
+		}
+
+		ret = do_drive_set_taskfiles(ap, &ap->device[ix],
+				gtf_length, gtf_address);
+		kfree((void *)obj_loc);
+		if (ret < 0) {
+			if (ata_msg_probe(ap))
+				ata_port_printk(ap, KERN_DEBUG,
+					"%s: set_taskfiles error (%d)\n",
+					__FUNCTION__, ret);
+			break;
+		}
+	}
+
+	return ret;
+}
+
+/**
+ * ata_acpi_push_id - send Identify data to drive
+ * @ap: the ata_port for the drive
+ * @ix: drive index
+ *
+ * _SDD ACPI object: for SATA mode only
+ * Must be after Identify (Packet) Device -- uses its data
+ * ATM this function never returns a failure.  It is an optional
+ * method and if it fails for whatever reason, we should still
+ * just keep going.
+ */
+int ata_acpi_push_id(struct ata_port *ap, unsigned int ix)
+{
+	acpi_handle                     handle;
+	acpi_integer                    pcidevfn;
+	int                             err;
+	struct device                   *dev = ap->host->dev;
+	struct ata_device               *atadev = &ap->device[ix];
+	u32                             dev_adr;
+	acpi_status                     status;
+	struct acpi_object_list         input;
+	union acpi_object               in_params[1];
+
+	if (noacpi)
+		return 0;
+
+	if (ata_msg_probe(ap))
+		ata_dev_printk(atadev, KERN_DEBUG,
+			"%s: ap->id: %d, ix = %d, port#: %d\n",
+			__FUNCTION__, ap->id, ix, ap->port_no);
+
+	/* Don't continue if not a SATA device. */
+	if (!(ap->cbl == ATA_CBL_SATA)) {
+		if (ata_msg_probe(ap))
+			ata_dev_printk(atadev, KERN_DEBUG,
+				"%s: Not a SATA device\n", __FUNCTION__);
+		goto out;
+	}
+
+	/* Don't continue if device has no _ADR method.
+	 * _SDD is intended for known motherboard devices. */
+	err = sata_get_dev_handle(dev, &handle, &pcidevfn);
+	if (err < 0) {
+		if (ata_msg_probe(ap))
+			ata_dev_printk(atadev, KERN_DEBUG,
+				"%s: sata_get_dev_handle failed (%d\n",
+				__FUNCTION__, err);
+		goto out;
+	}
+
+	/* Get this drive's _ADR info, if not already known */
+	if (!atadev->obj_handle) {
+		dev_adr = SATA_ADR_RSVD;
+		err = get_sata_adr(dev, handle, pcidevfn, ix, ap, atadev,
+					&dev_adr);
+		if (err < 0 || dev_adr == SATA_ADR_RSVD ||
+			!atadev->obj_handle) {
+			if (ata_msg_probe(ap))
+				ata_dev_printk(atadev, KERN_DEBUG,
+					"%s: get_sata_adr failed: "
+					"err=%d, dev_adr=%u, obj_handle=0x%p\n",
+					__FUNCTION__, err, dev_adr,
+					atadev->obj_handle);
+			goto out;
+		}
+	}
+
+	/* Give the drive Identify data to the drive via the _SDD method */
+	/* _SDD: set up input parameters */
+	input.count = 1;
+	input.pointer = in_params;
+	in_params[0].type = ACPI_TYPE_BUFFER;
+	in_params[0].buffer.length = sizeof(atadev->id[0]) * ATA_ID_WORDS;
+	in_params[0].buffer.pointer = (u8 *)atadev->id;
+	/* Output buffer: _SDD has no output */
+
+	/* It's OK for _SDD to be missing too. */
+	swap_buf_le16(atadev->id, ATA_ID_WORDS);
+	status = acpi_evaluate_object(atadev->obj_handle, "_SDD", &input, NULL);
+	swap_buf_le16(atadev->id, ATA_ID_WORDS);
+
+	err = ACPI_FAILURE(status) ? -EIO : 0;
+	if (err < 0) {
+		if (ata_msg_probe(ap))
+			ata_dev_printk(atadev, KERN_DEBUG,
+				"ata%u(%u): %s _SDD error: status = 0x%x\n",
+				ap->id, ap->device->devno,
+				__FUNCTION__, status);
+	}
+
+	/* always return success */
+out:
+	return 0;
+}
+
+
diff --git a/drivers/ata/libata-core.c b/drivers/ata/libata-core.c
index 2cf8251..e900c5e 100644
--- a/drivers/ata/libata-core.c
+++ b/drivers/ata/libata-core.c
@@ -93,6 +93,10 @@ static int ata_probe_timeout = ATA_TMOUT_INTERNAL / HZ;
 module_param(ata_probe_timeout, int, 0444);
 MODULE_PARM_DESC(ata_probe_timeout, "Set ATA probing timeout (seconds)");
 
+int noacpi;
+module_param(noacpi, int, 0444);
+MODULE_PARM_DESC(noacpi, "Disables the use of ACPI in suspend/resume when set");
+
 MODULE_AUTHOR("Jeff Garzik");
 MODULE_DESCRIPTION("Library module for ATA devices");
 MODULE_LICENSE("GPL");
@@ -1564,6 +1568,16 @@ int ata_dev_configure(struct ata_device *dev)
 		ata_dev_printk(dev, KERN_DEBUG, "%s: ENTER, host %u, dev %u\n",
 			       __FUNCTION__, ap->id, dev->devno);
 
+	/* set _SDD */
+	rc = ata_acpi_push_id(ap, dev->devno);
+	if (rc) {
+		ata_dev_printk(dev, KERN_WARNING, "failed to set _SDD(%d)\n",
+			rc);
+	}
+
+	/* retrieve and execute the ATA task file of _GTF */
+	ata_acpi_exec_tfs(ap);
+
 	/* print device capabilities */
 	if (ata_msg_probe(ap))
 		ata_dev_printk(dev, KERN_DEBUG,
diff --git a/drivers/ata/libata.h b/drivers/ata/libata.h
index 06ccf23..0ad7781 100644
--- a/drivers/ata/libata.h
+++ b/drivers/ata/libata.h
@@ -47,6 +47,7 @@ extern struct workqueue_struct *ata_aux_wq;
 extern int atapi_enabled;
 extern int atapi_dmadir;
 extern int libata_fua;
+extern int noacpi;
 extern struct ata_queued_cmd *ata_qc_new_init(struct ata_device *dev);
 extern int ata_build_rw_tf(struct ata_taskfile *tf, struct ata_device *dev,
 			   u64 block, u32 n_block, unsigned int tf_flags,
@@ -87,6 +88,20 @@ extern void ata_port_init(struct ata_port *ap, struct ata_host *host,
 extern struct ata_probe_ent *ata_probe_ent_alloc(struct device *dev,
 						 const struct ata_port_info *port);
 
+/* libata-acpi.c */
+#ifdef CONFIG_SATA_ACPI
+extern int ata_acpi_exec_tfs(struct ata_port *ap);
+extern int ata_acpi_push_id(struct ata_port *ap, unsigned int ix);
+#else
+static inline int ata_acpi_exec_tfs(struct ata_port *ap)
+{
+	return 0;
+}
+static inline int ata_acpi_push_id(struct ata_port *ap, unsigned int ix)
+{
+	return 0;
+}
+#endif
 
 /* libata-scsi.c */
 extern struct scsi_transport_template ata_scsi_transport_template;
diff --git a/include/linux/libata.h b/include/linux/libata.h
index b870b20..86762a9 100644
--- a/include/linux/libata.h
+++ b/include/linux/libata.h
@@ -35,6 +35,7 @@
 #include <linux/ata.h>
 #include <linux/workqueue.h>
 #include <scsi/scsi_host.h>
+#include <linux/acpi.h>
 
 /*
  * Define if arch has non-standard setup.  This is a _PCI_ standard
@@ -495,6 +496,10 @@ struct ata_device {
 	/* error history */
 	struct ata_ering	ering;
 	unsigned int		horkage;	/* List of broken features */
+#ifdef CONFIG_SATA_ACPI
+	/* ACPI objects info */
+	acpi_handle obj_handle;
+#endif
 };
 
 /* Offset into struct ata_device.  Fields above it are maintained

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [git patches] libata ACPI support
  2007-02-20 23:49     ` Jeff Garzik
  2007-02-21 11:39       ` Lukas Hejtmanek
@ 2007-02-21 17:56       ` Lukas Hejtmanek
  1 sibling, 0 replies; 10+ messages in thread
From: Lukas Hejtmanek @ 2007-02-21 17:56 UTC (permalink / raw)
  To: Jeff Garzik; +Cc: linux-kernel, IDE/ATA development list

On Tue, Feb 20, 2007 at 06:49:17PM -0500, Jeff Garzik wrote:
> Can you narrow down the -gitX release further?  i.e. does 2.6.20-git13 
> perform at full speed?  -git13 includes several libata changes, but does 
> not include the ACPI additions.
> 
> Since this is an ACPI execution error in the bug, that would tend to 
> point to bad BIOS tables.  But since there is a performance drop that 
> does not disappear with acpi=off, there might also be another factor, 
> unrelated to ACPI, at play here.

I've noticed on lkm another bug report which has pointed me to this:
ata_piix 0000:00:1f.1: version 2.00ac7
ACPI: PCI Interrupt 0000:00:1f.1[A] -> GSI 18 (level, low) -> IRQ 18
PCI: Setting latency timer of device 0000:00:1f.1 to 64
ata1: PATA max UDMA/100 cmd 0x000101f0 ctl 0x000103f6 bmdma 0x0001ffa0 irq 14
ata2: PATA max UDMA/100 cmd 0x00010170 ctl 0x00010376 bmdma 0x0001ffa8 irq 15
scsi0 : ata_piix
ata1.00: ATA-6: ST9100824A, 3.04, max UDMA/100
ata1.00: 195371568 sectors, multi 16: LBA48
ata1.01: ATAPI, max UDMA/33
ata1.00: configured for UDMA/33
ata1.01: configured for UDMA/33

Both ports are for some reason configured as UDMA/33.

-- 
Lukáš Hejtmánek

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [git patches] libata ACPI support
  2007-02-20 23:49     ` Jeff Garzik
@ 2007-02-21 11:39       ` Lukas Hejtmanek
  2007-02-21 17:56       ` Lukas Hejtmanek
  1 sibling, 0 replies; 10+ messages in thread
From: Lukas Hejtmanek @ 2007-02-21 11:39 UTC (permalink / raw)
  To: Jeff Garzik; +Cc: linux-kernel, IDE/ATA development list

On Tue, Feb 20, 2007 at 06:49:17PM -0500, Jeff Garzik wrote:
> Can you narrow down the -gitX release further?  i.e. does 2.6.20-git13 
> perform at full speed?  -git13 includes several libata changes, but does 
> not include the ACPI additions.
> 
> Since this is an ACPI execution error in the bug, that would tend to 
> point to bad BIOS tables.  But since there is a performance drop that 
> does not disappear with acpi=off, there might also be another factor, 
> unrelated to ACPI, at play here.

I have found that the issue is between git12 and git13.

I have also found another issue, not sure whether it is also presented by
git13, however in git15, IDE interrupts are no more unmasked. 

-- 
Lukáš Hejtmánek

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [git patches] libata ACPI support
  2007-02-20 21:47 Lukas Hejtmanek
  2007-02-20 22:23 ` Jeff Garzik
  2007-02-20 22:23 ` Jeff Garzik
@ 2007-02-21  1:20 ` Alan
  2 siblings, 0 replies; 10+ messages in thread
From: Alan @ 2007-02-21  1:20 UTC (permalink / raw)
  To: Lukas Hejtmanek; +Cc: jgarzik, linux-kernel

On Tue, 20 Feb 2007 22:47:17 +0100
Lukas Hejtmanek <xhejtman@mail.muni.cz> wrote:

> Hello,
> 
> introducing SATA ACPI does not work for me (I've filled the bug
> report: http://bugzilla.kernel.org/show_bug.cgi?id=8046) beside those error
> messages, IDE performance is decreased. libata.noacpi=1 option works ok but
> the performance is still slightly degraded (about 10%).

The performance degradation appears unrelated to the ACPI errors. If you
ignore the ACPI errors something has broken ata_piix cable detection.

Alan

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [git patches] libata ACPI support
  2007-02-20 23:26   ` Lukas Hejtmanek
@ 2007-02-20 23:49     ` Jeff Garzik
  2007-02-21 11:39       ` Lukas Hejtmanek
  2007-02-21 17:56       ` Lukas Hejtmanek
  0 siblings, 2 replies; 10+ messages in thread
From: Jeff Garzik @ 2007-02-20 23:49 UTC (permalink / raw)
  To: Lukas Hejtmanek; +Cc: linux-kernel, IDE/ATA development list

Lukas Hejtmanek wrote:
> On Tue, Feb 20, 2007 at 05:23:21PM -0500, Jeff Garzik wrote:
>>> introducing SATA ACPI does not work for me (I've filled the bug
>>> report: http://bugzilla.kernel.org/show_bug.cgi?id=8046) beside those error
>>> messages, IDE performance is decreased. libata.noacpi=1 option works ok but
>>> the performance is still slightly degraded (about 10%).
>> Does performance come back if you also use acpi=off ?
> 
> It does not.
> 
> With 2.6.20-git15, hdparm -t /dev/sda says 28MB/s, 
> with 2.6.20, hdparm -t /dev/sda says 32MB/s.

Can you narrow down the -gitX release further?  i.e. does 2.6.20-git13 
perform at full speed?  -git13 includes several libata changes, but does 
not include the ACPI additions.

Since this is an ACPI execution error in the bug, that would tend to 
point to bad BIOS tables.  But since there is a performance drop that 
does not disappear with acpi=off, there might also be another factor, 
unrelated to ACPI, at play here.

	Jeff




^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [git patches] libata ACPI support
  2007-02-20 22:23 ` Jeff Garzik
@ 2007-02-20 23:26   ` Lukas Hejtmanek
  2007-02-20 23:49     ` Jeff Garzik
  0 siblings, 1 reply; 10+ messages in thread
From: Lukas Hejtmanek @ 2007-02-20 23:26 UTC (permalink / raw)
  To: Jeff Garzik; +Cc: linux-kernel

On Tue, Feb 20, 2007 at 05:23:21PM -0500, Jeff Garzik wrote:
> >introducing SATA ACPI does not work for me (I've filled the bug
> >report: http://bugzilla.kernel.org/show_bug.cgi?id=8046) beside those error
> >messages, IDE performance is decreased. libata.noacpi=1 option works ok but
> >the performance is still slightly degraded (about 10%).
> 
> Does performance come back if you also use acpi=off ?

It does not.

With 2.6.20-git15, hdparm -t /dev/sda says 28MB/s, 
with 2.6.20, hdparm -t /dev/sda says 32MB/s.

-- 
Lukáš Hejtmánek

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [git patches] libata ACPI support
  2007-02-20 22:23 ` Jeff Garzik
@ 2007-02-20 23:15   ` Lukas Hejtmanek
  0 siblings, 0 replies; 10+ messages in thread
From: Lukas Hejtmanek @ 2007-02-20 23:15 UTC (permalink / raw)
  To: Jeff Garzik; +Cc: linux-kernel

On Tue, Feb 20, 2007 at 05:23:56PM -0500, Jeff Garzik wrote:
> >introducing SATA ACPI does not work for me (I've filled the bug
> >report: http://bugzilla.kernel.org/show_bug.cgi?id=8046) beside those error
> >messages, IDE performance is decreased. libata.noacpi=1 option works ok but
> >the performance is still slightly degraded (about 10%).
> 
> Also, please add hardware info (full dmesg, full lspci) to the bug.

done.

-- 
Lukáš Hejtmánek

^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [git patches] libata ACPI support
  2007-02-20 21:47 Lukas Hejtmanek
  2007-02-20 22:23 ` Jeff Garzik
@ 2007-02-20 22:23 ` Jeff Garzik
  2007-02-20 23:15   ` Lukas Hejtmanek
  2007-02-21  1:20 ` Alan
  2 siblings, 1 reply; 10+ messages in thread
From: Jeff Garzik @ 2007-02-20 22:23 UTC (permalink / raw)
  To: Lukas Hejtmanek; +Cc: linux-kernel

Lukas Hejtmanek wrote:
> Hello,
> 
> introducing SATA ACPI does not work for me (I've filled the bug
> report: http://bugzilla.kernel.org/show_bug.cgi?id=8046) beside those error
> messages, IDE performance is decreased. libata.noacpi=1 option works ok but
> the performance is still slightly degraded (about 10%).

Also, please add hardware info (full dmesg, full lspci) to the bug.

	Jeff




^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [git patches] libata ACPI support
  2007-02-20 21:47 Lukas Hejtmanek
@ 2007-02-20 22:23 ` Jeff Garzik
  2007-02-20 23:26   ` Lukas Hejtmanek
  2007-02-20 22:23 ` Jeff Garzik
  2007-02-21  1:20 ` Alan
  2 siblings, 1 reply; 10+ messages in thread
From: Jeff Garzik @ 2007-02-20 22:23 UTC (permalink / raw)
  To: Lukas Hejtmanek; +Cc: linux-kernel

Lukas Hejtmanek wrote:
> Hello,
> 
> introducing SATA ACPI does not work for me (I've filled the bug
> report: http://bugzilla.kernel.org/show_bug.cgi?id=8046) beside those error
> messages, IDE performance is decreased. libata.noacpi=1 option works ok but
> the performance is still slightly degraded (about 10%).

Does performance come back if you also use acpi=off ?

	Jeff




^ permalink raw reply	[flat|nested] 10+ messages in thread

* Re: [git patches] libata ACPI support
@ 2007-02-20 21:47 Lukas Hejtmanek
  2007-02-20 22:23 ` Jeff Garzik
                   ` (2 more replies)
  0 siblings, 3 replies; 10+ messages in thread
From: Lukas Hejtmanek @ 2007-02-20 21:47 UTC (permalink / raw)
  To: jgarzik, linux-kernel

Hello,

introducing SATA ACPI does not work for me (I've filled the bug
report: http://bugzilla.kernel.org/show_bug.cgi?id=8046) beside those error
messages, IDE performance is decreased. libata.noacpi=1 option works ok but
the performance is still slightly degraded (about 10%).

-- 
Lukáš Hejtmánek

^ permalink raw reply	[flat|nested] 10+ messages in thread

end of thread, other threads:[~2007-02-21 17:56 UTC | newest]

Thread overview: 10+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2007-02-17 22:12 [git patches] libata ACPI support Jeff Garzik
2007-02-20 21:47 Lukas Hejtmanek
2007-02-20 22:23 ` Jeff Garzik
2007-02-20 23:26   ` Lukas Hejtmanek
2007-02-20 23:49     ` Jeff Garzik
2007-02-21 11:39       ` Lukas Hejtmanek
2007-02-21 17:56       ` Lukas Hejtmanek
2007-02-20 22:23 ` Jeff Garzik
2007-02-20 23:15   ` Lukas Hejtmanek
2007-02-21  1:20 ` Alan

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).