linux-scsi.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
* [RESEND][PATCH 06/10][SCSI]mpt2sas: For >2TB volumes, DirectDrive support sends IO's with LBA bit 31 to IR FW instead of DirectDrive
@ 2014-06-25 10:34 Reddy, Sreekanth
  2014-07-13 15:27 ` Martin K. Petersen
  0 siblings, 1 reply; 6+ messages in thread
From: Reddy, Sreekanth @ 2014-06-25 10:34 UTC (permalink / raw)
  To: jejb, JBottomley
  Cc: linux-scsi, Sathya.Prakash, Nagalakshmi.Nandigama,
	sreekanth.reddy, linux-kernel, hch, martin.petersen

There was a down casting of the volume max LBA from a U64 to a U32,
which is taken out and now the max LBA is set appropriately to U64.

Signed-off-by: Sreekanth Reddy <Sreekanth.Reddy@avagotech.com>
---
 drivers/scsi/mpt2sas/mpt2sas_scsih.c |   14 ++++++++++----
 1 files changed, 10 insertions(+), 4 deletions(-)

diff --git a/drivers/scsi/mpt2sas/mpt2sas_scsih.c b/drivers/scsi/mpt2sas/mpt2sas_scsih.c
index 6ae109b..4a0728a 100644
--- a/drivers/scsi/mpt2sas/mpt2sas_scsih.c
+++ b/drivers/scsi/mpt2sas/mpt2sas_scsih.c
@@ -3865,7 +3865,8 @@ _scsih_setup_direct_io(struct MPT2SAS_ADAPTER *ioc, struct scsi_cmnd *scmd,
 	struct _raid_device *raid_device, Mpi2SCSIIORequest_t *mpi_request,
 	u16 smid)
 {
-	u32 v_lba, p_lba, stripe_off, stripe_unit, column, io_size;
+	u32 p_lba, stripe_off, stripe_unit, column, io_size;
+	u64 v_lba;
 	u32 stripe_sz, stripe_exp;
 	u8 num_pds, *cdb_ptr, i;
 	u8 cdb0 = scmd->cmnd[0];
@@ -3882,12 +3883,17 @@ _scsih_setup_direct_io(struct MPT2SAS_ADAPTER *ioc, struct scsi_cmnd *scmd,
 			| cdb_ptr[5])) {
 			io_size = scsi_bufflen(scmd) >>
 			    raid_device->block_exponent;
-			i = (cdb0 < READ_16) ? 2 : 6;
+
 			/* get virtual lba */
-			v_lba = be32_to_cpu(*(__be32 *)(&cdb_ptr[i]));
+			if (cdb0 < READ_16)
+				v_lba = be32_to_cpu(*(__be32 *)(&cdb_ptr[2]));
+			else
+				v_lba = be64_to_cpu(*(__be64 *)(&cdb_ptr[2]));
+
+			i = (cdb0 < READ_16) ? 2 : 6;
 
 			if (((u64)v_lba + (u64)io_size - 1) <=
-			    (u32)raid_device->max_lba) {
+				raid_device->max_lba) {
 				stripe_sz = raid_device->stripe_sz;
 				stripe_exp = raid_device->stripe_exponent;
 				stripe_off = v_lba & (stripe_sz - 1);
-- 
1.7.1

^ permalink raw reply related	[flat|nested] 6+ messages in thread

* Re: [RESEND][PATCH 06/10][SCSI]mpt2sas: For >2TB volumes, DirectDrive support sends IO's with LBA bit 31 to IR FW instead of DirectDrive
  2014-06-25 10:34 [RESEND][PATCH 06/10][SCSI]mpt2sas: For >2TB volumes, DirectDrive support sends IO's with LBA bit 31 to IR FW instead of DirectDrive Reddy, Sreekanth
@ 2014-07-13 15:27 ` Martin K. Petersen
  2014-07-13 17:02   ` Elliott, Robert (Server Storage)
       [not found]   ` <CAK=zhgosXeox35kVvuiTm8EZaD2-e_Lxz79uFFAP0c1cMJNbcg@mail.gmail.com>
  0 siblings, 2 replies; 6+ messages in thread
From: Martin K. Petersen @ 2014-07-13 15:27 UTC (permalink / raw)
  To: Reddy, Sreekanth
  Cc: jejb, JBottomley, linux-scsi, Sathya.Prakash,
	Nagalakshmi.Nandigama, linux-kernel, hch, martin.petersen

>>>>> "Sreekanth" == Reddy, Sreekanth <Sreekanth.Reddy@avagotech.com> writes:

diff --git a/drivers/scsi/mpt2sas/mpt2sas_scsih.c b/drivers/scsi/mpt2sas/mpt2sas_scsih.c
index 6ae109b..4a0728a 100644
--- a/drivers/scsi/mpt2sas/mpt2sas_scsih.c
+++ b/drivers/scsi/mpt2sas/mpt2sas_scsih.c
@@ -3865,7 +3865,8 @@ _scsih_setup_direct_io(struct MPT2SAS_ADAPTER *ioc, struct scsi_cmnd *scmd,
 	struct _raid_device *raid_device, Mpi2SCSIIORequest_t *mpi_request,
 	u16 smid)
 {
-	u32 v_lba, p_lba, stripe_off, stripe_unit, column, io_size;
+	u32 p_lba, stripe_off, stripe_unit, column, io_size;
+	u64 v_lba;
 	u32 stripe_sz, stripe_exp;
 	u8 num_pds, *cdb_ptr, i;
 	u8 cdb0 = scmd->cmnd[0];
@@ -3882,12 +3883,17 @@ _scsih_setup_direct_io(struct MPT2SAS_ADAPTER *ioc, struct scsi_cmnd *scmd,
 			| cdb_ptr[5])) {
 			io_size = scsi_bufflen(scmd) >>
 			    raid_device->block_exponent;
-			i = (cdb0 < READ_16) ? 2 : 6;
+
 			/* get virtual lba */
-			v_lba = be32_to_cpu(*(__be32 *)(&cdb_ptr[i]));
+			if (cdb0 < READ_16)
+				v_lba = be32_to_cpu(*(__be32 *)(&cdb_ptr[2]));
+			else
+				v_lba = be64_to_cpu(*(__be64 *)(&cdb_ptr[2]));

Why aren't you using scsi_get_lba() instead of all this nasty CDB
parsing?

+
+			i = (cdb0 < READ_16) ? 2 : 6;

What about WRITE_16?  WRITE_16 > READ_16.

 			if (((u64)v_lba + (u64)io_size - 1) <=
-			    (u32)raid_device->max_lba) {
+				raid_device->max_lba) {
 				stripe_sz = raid_device->stripe_sz;
 				stripe_exp = raid_device->stripe_exponent;
 				stripe_off = v_lba & (stripe_sz - 1);

Also, this is not touched by the patch, but you're then doing:

        (*(__be32 *)(&cdb_ptr[i])) = cpu_to_be32(p_lba);

What if this is a 6-byte READ/WRITE command? You'll end up exceeding the
size of the LBA field.

What if you're using a 16-byte CDB and the target device LBA is > 2TB?

-- 
Martin K. Petersen	Oracle Linux Engineering

^ permalink raw reply related	[flat|nested] 6+ messages in thread

* RE: [RESEND][PATCH 06/10][SCSI]mpt2sas: For >2TB volumes, DirectDrive support sends IO's with LBA bit 31 to IR FW instead of DirectDrive
  2014-07-13 15:27 ` Martin K. Petersen
@ 2014-07-13 17:02   ` Elliott, Robert (Server Storage)
       [not found]   ` <CAK=zhgosXeox35kVvuiTm8EZaD2-e_Lxz79uFFAP0c1cMJNbcg@mail.gmail.com>
  1 sibling, 0 replies; 6+ messages in thread
From: Elliott, Robert (Server Storage) @ 2014-07-13 17:02 UTC (permalink / raw)
  To: Martin K. Petersen, Reddy, Sreekanth
  Cc: jejb@kernel.org, JBottomley@Parallels.com,
	linux-scsi@vger.kernel.org, Sathya.Prakash@avagotech.com,
	Nagalakshmi.Nandigama@avagotech.com, linux-kernel@vger.kernel.org,
	hch@infradead.org



> -----Original Message-----
> From: linux-scsi-owner@vger.kernel.org [mailto:linux-scsi-
> owner@vger.kernel.org] On Behalf Of Martin K. Petersen


...
> Also, this is not touched by the patch, but you're then doing:
> 
>         (*(__be32 *)(&cdb_ptr[i])) = cpu_to_be32(p_lba);
> 
> What if this is a 6-byte READ/WRITE command? You'll end up exceeding the
> size of the LBA field.

All this is inside:
        if (cdb0 == READ_16 || cdb0 == READ_10 ||
            cdb0 == WRITE_16 || cdb0 == WRITE_10) {

so READ_6 and WRITE_6 and all their oddities are not a problem here.




^ permalink raw reply	[flat|nested] 6+ messages in thread

* Re: [RESEND][PATCH 06/10][SCSI]mpt2sas: For >2TB volumes, DirectDrive support sends IO's with LBA bit 31 to IR FW instead of DirectDrive
       [not found]         ` <CAK=zhgrUWOS-pPjamF+z+AZZvdXz_JxyKgdTv6Zae0oLk=qQog@mail.gmail.com>
@ 2014-07-15 14:00           ` Martin K. Petersen
       [not found]             ` <CAK=zhgrE0QBgyyLTU0Jdq2cAbKbd=wY4zfMGo2ux9nWz9oBgiQ@mail.gmail.com>
  0 siblings, 1 reply; 6+ messages in thread
From: Martin K. Petersen @ 2014-07-15 14:00 UTC (permalink / raw)
  To: Sreekanth Reddy
  Cc: Martin K. Petersen, Nagalakshmi Nandigama, Sathya Prakash,
	linux-scsi

>>>>> "Sreekanth" == Sreekanth Reddy <sreekanth.reddy@avagotech.com> writes:

Sreekanth> Also coming to first point i.e. the usage of scsi_get_lba()
Sreekanth> API instead of parsing the CDB. During an experiment I
Sreekanth> observe that when I use the 'sg_dd if=/dev/sg0 of=/dev/null
Sreekanth> count=1' command for read operation then the output of this
Sreekanth> API is all FFFF's i.e.  0xffffffff (i.e. LBA is 0xffffffff
Sreekanth> instead of 0x00000000 for 32 bit LBA).

scsi_get_lba() returns a sector_t. I suspect it is all the type casting
that is messing things up.

>> p_lba is 32-bit but READ(16)/WRITE(16) take an 8-byte LBA. Your patch
>> addresses the v_lba being 64-bit but not p_lba.

I now see that you have two almost identical code paths to deal with
p_lba.

How about something like this (untested):

diff --git a/drivers/scsi/mpt2sas/mpt2sas_scsih.c b/drivers/scsi/mpt2sas/mpt2sas_scsih.c
index 5055f925d2cd..24efc83d8e11 100644
--- a/drivers/scsi/mpt2sas/mpt2sas_scsih.c
+++ b/drivers/scsi/mpt2sas/mpt2sas_scsih.c
@@ -55,6 +55,8 @@
 #include <linux/raid_class.h>
 #include <linux/slab.h>
 
+#include <asm/unaligned.h>
+
 #include "mpt2sas_base.h"
 
 MODULE_AUTHOR(MPT2SAS_AUTHOR);
@@ -3857,85 +3859,41 @@ _scsih_setup_direct_io(struct MPT2SAS_ADAPTER *ioc, struct scsi_cmnd *scmd,
 	struct _raid_device *raid_device, Mpi2SCSIIORequest_t *mpi_request,
 	u16 smid)
 {
-	u32 v_lba, p_lba, stripe_off, stripe_unit, column, io_size;
+	sector_t v_lba, p_lba, stripe_off, stripe_unit, column, io_size;
 	u32 stripe_sz, stripe_exp;
-	u8 num_pds, *cdb_ptr, i;
-	u8 cdb0 = scmd->cmnd[0];
-	u64 v_llba;
+	u8 num_pds, cmd = scmd->cmnd[0];
 
-	/*
-	 * Try Direct I/O to RAID memeber disks
-	 */
-	if (cdb0 == READ_16 || cdb0 == READ_10 ||
-	    cdb0 == WRITE_16 || cdb0 == WRITE_10) {
-		cdb_ptr = mpi_request->CDB.CDB32;
-
-		if ((cdb0 < READ_16) || !(cdb_ptr[2] | cdb_ptr[3] | cdb_ptr[4]
-			| cdb_ptr[5])) {
-			io_size = scsi_bufflen(scmd) >>
-			    raid_device->block_exponent;
-			i = (cdb0 < READ_16) ? 2 : 6;
-			/* get virtual lba */
-			v_lba = be32_to_cpu(*(__be32 *)(&cdb_ptr[i]));
-
-			if (((u64)v_lba + (u64)io_size - 1) <=
-			    (u32)raid_device->max_lba) {
-				stripe_sz = raid_device->stripe_sz;
-				stripe_exp = raid_device->stripe_exponent;
-				stripe_off = v_lba & (stripe_sz - 1);
-
-				/* Check whether IO falls within a stripe */
-				if ((stripe_off + io_size) <= stripe_sz) {
-					num_pds = raid_device->num_pds;
-					p_lba = v_lba >> stripe_exp;
-					stripe_unit = p_lba / num_pds;
-					column = p_lba % num_pds;
-					p_lba = (stripe_unit << stripe_exp) +
-					    stripe_off;
-					mpi_request->DevHandle =
-						cpu_to_le16(raid_device->
-						    pd_handle[column]);
-					(*(__be32 *)(&cdb_ptr[i])) =
-						cpu_to_be32(p_lba);
-					/*
-					* WD: To indicate this I/O is directI/O
-					*/
-					_scsih_scsi_direct_io_set(ioc, smid, 1);
-				}
-			}
-		} else {
-			io_size = scsi_bufflen(scmd) >>
-			    raid_device->block_exponent;
-			/* get virtual lba */
-			v_llba = be64_to_cpu(*(__be64 *)(&cdb_ptr[2]));
-
-			if ((v_llba + (u64)io_size - 1) <=
-			    raid_device->max_lba) {
-				stripe_sz = raid_device->stripe_sz;
-				stripe_exp = raid_device->stripe_exponent;
-				stripe_off = (u32) (v_llba & (stripe_sz - 1));
-
-				/* Check whether IO falls within a stripe */
-				if ((stripe_off + io_size) <= stripe_sz) {
-					num_pds = raid_device->num_pds;
-					p_lba = (u32)(v_llba >> stripe_exp);
-					stripe_unit = p_lba / num_pds;
-					column = p_lba % num_pds;
-					p_lba = (stripe_unit << stripe_exp) +
-					    stripe_off;
-					mpi_request->DevHandle =
-						cpu_to_le16(raid_device->
-						    pd_handle[column]);
-					(*(__be64 *)(&cdb_ptr[2])) =
-					    cpu_to_be64((u64)p_lba);
-					/*
-					* WD: To indicate this I/O is directI/O
-					*/
-					_scsih_scsi_direct_io_set(ioc, smid, 1);
-				}
-			}
-		}
-	}
+	if (cmd != READ_10 && cmd != WRITE_10 &&
+	    cmd != READ_16 && cmd != WRITE_16)
+		return;
+
+	v_lba = scsi_get_lba(scmd);
+	io_size = scsi_bufflen(scmd) >> raid_device->block_exponent;
+
+	if (v_lba + io_size - 1 > raid_device->max_lba)
+		return;
+
+	stripe_sz = raid_device->stripe_sz;
+	stripe_exp = raid_device->stripe_exponent;
+	stripe_off = v_lba & (stripe_sz - 1);
+
+	/* Return unless IO falls within a stripe */
+	if (stripe_off + io_size > stripe_sz)
+		return;
+
+	num_pds = raid_device->num_pds;
+	p_lba = v_lba >> stripe_exp;
+	stripe_unit = p_lba / num_pds;
+	column = p_lba % num_pds;
+	p_lba = (stripe_unit << stripe_exp) + stripe_off;
+	mpi_request->DevHandle = cpu_to_le16(raid_device->pd_handle[column]);
+
+	if (cmd == READ_10 || cmd == WRITE_10)
+		put_unaligned_be32(lower_32_bits(p_lba), mpi_request->CDB.CDB32);
+	else
+		put_unaligned_be64(p_lba, mpi_request->CDB.CDB32);
+
+	_scsih_scsi_direct_io_set(ioc, smid, 1);
 }
 
 /**

^ permalink raw reply related	[flat|nested] 6+ messages in thread

* Re: [RESEND][PATCH 06/10][SCSI]mpt2sas: For >2TB volumes, DirectDrive support sends IO's with LBA bit 31 to IR FW instead of DirectDrive
       [not found]               ` <CAK=zhgpqcFu5WKLEOD9E=K9DURO9HEtXE8VDVpScLiZK+3JsAQ@mail.gmail.com>
@ 2014-07-15 16:02                 ` Martin K. Petersen
       [not found]                   ` <CAK=zhgoNm=STXjBOOqXHaz4yPuRHQ0avwgUOD_H0B==sgt6vEg@mail.gmail.com>
  0 siblings, 1 reply; 6+ messages in thread
From: Martin K. Petersen @ 2014-07-15 16:02 UTC (permalink / raw)
  To: Sreekanth Reddy
  Cc: Martin K. Petersen, Nagalakshmi Nandigama, Sathya Prakash,
	linux-scsi

>>>>> "Sreekanth" == Sreekanth Reddy <sreekanth.reddy@avagotech.com> writes:

Sreekanth> I have tried with this diff. And I still observe the invalid
Sreekanth> LBA (i.e.  0xFFFFFFFF) with scsi_get_lba() API when I used
Sreekanth> the sg_dd with input file specified as /dev/sg0.

Yeah, I now remember that scsi_get_lba() fails for PC commands.

Sreekanth> Also below section of code will overwrite the CDB operation
Sreekanth> code. and the final CDB has the some junk values.

Yep, already fixed that.

Please try the updated patch below.

diff --git a/drivers/scsi/mpt2sas/mpt2sas_scsih.c b/drivers/scsi/mpt2sas/mpt2sas_scsih.c
index 5055f925d2cd..11617f6b8c60 100644
--- a/drivers/scsi/mpt2sas/mpt2sas_scsih.c
+++ b/drivers/scsi/mpt2sas/mpt2sas_scsih.c
@@ -55,6 +55,8 @@
 #include <linux/raid_class.h>
 #include <linux/slab.h>
 
+#include <asm/unaligned.h>
+
 #include "mpt2sas_base.h"
 
 MODULE_AUTHOR(MPT2SAS_AUTHOR);
@@ -3857,85 +3859,47 @@ _scsih_setup_direct_io(struct MPT2SAS_ADAPTER *ioc, struct scsi_cmnd *scmd,
 	struct _raid_device *raid_device, Mpi2SCSIIORequest_t *mpi_request,
 	u16 smid)
 {
-	u32 v_lba, p_lba, stripe_off, stripe_unit, column, io_size;
+	sector_t v_lba, p_lba, stripe_off, stripe_unit, column, io_size;
 	u32 stripe_sz, stripe_exp;
-	u8 num_pds, *cdb_ptr, i;
-	u8 cdb0 = scmd->cmnd[0];
-	u64 v_llba;
+	u8 num_pds, cmd = scmd->cmnd[0];
 
-	/*
-	 * Try Direct I/O to RAID memeber disks
-	 */
-	if (cdb0 == READ_16 || cdb0 == READ_10 ||
-	    cdb0 == WRITE_16 || cdb0 == WRITE_10) {
-		cdb_ptr = mpi_request->CDB.CDB32;
-
-		if ((cdb0 < READ_16) || !(cdb_ptr[2] | cdb_ptr[3] | cdb_ptr[4]
-			| cdb_ptr[5])) {
-			io_size = scsi_bufflen(scmd) >>
-			    raid_device->block_exponent;
-			i = (cdb0 < READ_16) ? 2 : 6;
-			/* get virtual lba */
-			v_lba = be32_to_cpu(*(__be32 *)(&cdb_ptr[i]));
-
-			if (((u64)v_lba + (u64)io_size - 1) <=
-			    (u32)raid_device->max_lba) {
-				stripe_sz = raid_device->stripe_sz;
-				stripe_exp = raid_device->stripe_exponent;
-				stripe_off = v_lba & (stripe_sz - 1);
-
-				/* Check whether IO falls within a stripe */
-				if ((stripe_off + io_size) <= stripe_sz) {
-					num_pds = raid_device->num_pds;
-					p_lba = v_lba >> stripe_exp;
-					stripe_unit = p_lba / num_pds;
-					column = p_lba % num_pds;
-					p_lba = (stripe_unit << stripe_exp) +
-					    stripe_off;
-					mpi_request->DevHandle =
-						cpu_to_le16(raid_device->
-						    pd_handle[column]);
-					(*(__be32 *)(&cdb_ptr[i])) =
-						cpu_to_be32(p_lba);
-					/*
-					* WD: To indicate this I/O is directI/O
-					*/
-					_scsih_scsi_direct_io_set(ioc, smid, 1);
-				}
-			}
-		} else {
-			io_size = scsi_bufflen(scmd) >>
-			    raid_device->block_exponent;
-			/* get virtual lba */
-			v_llba = be64_to_cpu(*(__be64 *)(&cdb_ptr[2]));
-
-			if ((v_llba + (u64)io_size - 1) <=
-			    raid_device->max_lba) {
-				stripe_sz = raid_device->stripe_sz;
-				stripe_exp = raid_device->stripe_exponent;
-				stripe_off = (u32) (v_llba & (stripe_sz - 1));
-
-				/* Check whether IO falls within a stripe */
-				if ((stripe_off + io_size) <= stripe_sz) {
-					num_pds = raid_device->num_pds;
-					p_lba = (u32)(v_llba >> stripe_exp);
-					stripe_unit = p_lba / num_pds;
-					column = p_lba % num_pds;
-					p_lba = (stripe_unit << stripe_exp) +
-					    stripe_off;
-					mpi_request->DevHandle =
-						cpu_to_le16(raid_device->
-						    pd_handle[column]);
-					(*(__be64 *)(&cdb_ptr[2])) =
-					    cpu_to_be64((u64)p_lba);
-					/*
-					* WD: To indicate this I/O is directI/O
-					*/
-					_scsih_scsi_direct_io_set(ioc, smid, 1);
-				}
-			}
-		}
-	}
+	if (cmd != READ_10 && cmd != WRITE_10 &&
+	    cmd != READ_16 && cmd != WRITE_16)
+		return;
+
+	if (cmd == READ_10 || cmd == WRITE_10)
+		v_lba = get_unaligned_be32(&mpi_request->CDB.CDB32[2]);
+	else
+		v_lba = get_unaligned_be64(&mpi_request->CDB.CDB32[2]);
+
+	v_lba = scsi_get_lba(scmd);
+	io_size = scsi_bufflen(scmd) >> raid_device->block_exponent;
+
+	if (v_lba + io_size - 1 > raid_device->max_lba)
+		return;
+
+	stripe_sz = raid_device->stripe_sz;
+	stripe_exp = raid_device->stripe_exponent;
+	stripe_off = v_lba & (stripe_sz - 1);
+
+	/* Return unless IO falls within a stripe */
+	if (stripe_off + io_size > stripe_sz)
+		return;
+
+	num_pds = raid_device->num_pds;
+	p_lba = v_lba >> stripe_exp;
+	stripe_unit = p_lba / num_pds;
+	column = p_lba % num_pds;
+	p_lba = (stripe_unit << stripe_exp) + stripe_off;
+	mpi_request->DevHandle = cpu_to_le16(raid_device->pd_handle[column]);
+
+	if (cmd == READ_10 || cmd == WRITE_10)
+		put_unaligned_be32(lower_32_bits(p_lba),
+				   &mpi_request->CDB.CDB32[2]);
+	else
+		put_unaligned_be64(p_lba, &mpi_request->CDB.CDB32[2]);
+
+	_scsih_scsi_direct_io_set(ioc, smid, 1);
 }
 
 /**

^ permalink raw reply related	[flat|nested] 6+ messages in thread

* Re: [RESEND][PATCH 06/10][SCSI]mpt2sas: For >2TB volumes, DirectDrive support sends IO's with LBA bit 31 to IR FW instead of DirectDrive
       [not found]                   ` <CAK=zhgoNm=STXjBOOqXHaz4yPuRHQ0avwgUOD_H0B==sgt6vEg@mail.gmail.com>
@ 2014-07-15 17:02                     ` Martin K. Petersen
  0 siblings, 0 replies; 6+ messages in thread
From: Martin K. Petersen @ 2014-07-15 17:02 UTC (permalink / raw)
  To: Sreekanth Reddy
  Cc: Martin K. Petersen, Nagalakshmi Nandigama, Sathya Prakash,
	linux-scsi

>>>>> "Sreekanth" == Sreekanth Reddy <sreekanth.reddy@avagotech.com> writes:

Sreekanth> But still I observe the invalid LBA when we send IOs using
Sreekanth> sg_dd with input file specified as /dev/sg0 as shown below

In the latest patch I get the LBA from CDB32 in the MPI request so that
shouldn't happen.

-- 
Martin K. Petersen	Oracle Linux Engineering

^ permalink raw reply	[flat|nested] 6+ messages in thread

end of thread, other threads:[~2014-07-15 17:02 UTC | newest]

Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2014-06-25 10:34 [RESEND][PATCH 06/10][SCSI]mpt2sas: For >2TB volumes, DirectDrive support sends IO's with LBA bit 31 to IR FW instead of DirectDrive Reddy, Sreekanth
2014-07-13 15:27 ` Martin K. Petersen
2014-07-13 17:02   ` Elliott, Robert (Server Storage)
     [not found]   ` <CAK=zhgosXeox35kVvuiTm8EZaD2-e_Lxz79uFFAP0c1cMJNbcg@mail.gmail.com>
     [not found]     ` <CAK=zhgq6bXRBapfBy+Nx5uFFLUuj1cH78x0RXCU1Vk8Y9NWheA@mail.gmail.com>
     [not found]       ` <yq1fvi3y736.fsf@sermon.lab.mkp.net>
     [not found]         ` <CAK=zhgrUWOS-pPjamF+z+AZZvdXz_JxyKgdTv6Zae0oLk=qQog@mail.gmail.com>
2014-07-15 14:00           ` Martin K. Petersen
     [not found]             ` <CAK=zhgrE0QBgyyLTU0Jdq2cAbKbd=wY4zfMGo2ux9nWz9oBgiQ@mail.gmail.com>
     [not found]               ` <CAK=zhgpqcFu5WKLEOD9E=K9DURO9HEtXE8VDVpScLiZK+3JsAQ@mail.gmail.com>
2014-07-15 16:02                 ` Martin K. Petersen
     [not found]                   ` <CAK=zhgoNm=STXjBOOqXHaz4yPuRHQ0avwgUOD_H0B==sgt6vEg@mail.gmail.com>
2014-07-15 17:02                     ` Martin K. Petersen

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).