[PATCH v3 1/3] nvme/scsi: Remove power management support

2016-09-15 Thread Andy Lutomirski
As far as I can tell, there is basically nothing correct about this
code.  It misinterprets npss (off-by-one).  It hardcodes a bunch of
power states, which is nonsense, because they're all just indices
into a table that software needs to parse.  It completely ignores
the distinction between operational and non-operational states.
And, until 4.8, if all of the above magically succeeded, it would
dereference a NULL pointer and OOPS.

Since this code appears to be useless, just delete it.

Signed-off-by: Andy Lutomirski 
---
 drivers/nvme/host/scsi.c | 74 ++--
 1 file changed, 3 insertions(+), 71 deletions(-)

diff --git a/drivers/nvme/host/scsi.c b/drivers/nvme/host/scsi.c
index e947e298a737..44009105f8c8 100644
--- a/drivers/nvme/host/scsi.c
+++ b/drivers/nvme/host/scsi.c
@@ -72,15 +72,6 @@ static int sg_version_num = 30534;   /* 2 digits for each 
component */
 #define ALL_LUNS_RETURNED  0x02
 #define ALL_WELL_KNOWN_LUNS_RETURNED   0x01
 #define RESTRICTED_LUNS_RETURNED   0x00
-#define NVME_POWER_STATE_START_VALID   0x00
-#define NVME_POWER_STATE_ACTIVE0x01
-#define NVME_POWER_STATE_IDLE  0x02
-#define NVME_POWER_STATE_STANDBY   0x03
-#define NVME_POWER_STATE_LU_CONTROL0x07
-#define POWER_STATE_0  0
-#define POWER_STATE_1  1
-#define POWER_STATE_2  2
-#define POWER_STATE_3  3
 #define DOWNLOAD_SAVE_ACTIVATE 0x05
 #define DOWNLOAD_SAVE_DEFER_ACTIVATE   0x0E
 #define ACTIVATE_DEFERRED_MICROCODE0x0F
@@ -1229,64 +1220,6 @@ static void nvme_trans_fill_read_cap(u8 *response, 
struct nvme_id_ns *id_ns,
 
 /* Start Stop Unit Helper Functions */
 
-static int nvme_trans_power_state(struct nvme_ns *ns, struct sg_io_hdr *hdr,
-   u8 pc, u8 pcmod, u8 start)
-{
-   int res;
-   int nvme_sc;
-   struct nvme_id_ctrl *id_ctrl;
-   int lowest_pow_st;  /* max npss = lowest power consumption */
-   unsigned ps_desired = 0;
-
-   nvme_sc = nvme_identify_ctrl(ns->ctrl, &id_ctrl);
-   res = nvme_trans_status_code(hdr, nvme_sc);
-   if (res)
-   return res;
-
-   lowest_pow_st = max(POWER_STATE_0, (int)(id_ctrl->npss - 1));
-   kfree(id_ctrl);
-
-   switch (pc) {
-   case NVME_POWER_STATE_START_VALID:
-   /* Action unspecified if POWER CONDITION MODIFIER != 0 */
-   if (pcmod == 0 && start == 0x1)
-   ps_desired = POWER_STATE_0;
-   if (pcmod == 0 && start == 0x0)
-   ps_desired = lowest_pow_st;
-   break;
-   case NVME_POWER_STATE_ACTIVE:
-   /* Action unspecified if POWER CONDITION MODIFIER != 0 */
-   if (pcmod == 0)
-   ps_desired = POWER_STATE_0;
-   break;
-   case NVME_POWER_STATE_IDLE:
-   /* Action unspecified if POWER CONDITION MODIFIER != [0,1,2] */
-   if (pcmod == 0x0)
-   ps_desired = POWER_STATE_1;
-   else if (pcmod == 0x1)
-   ps_desired = POWER_STATE_2;
-   else if (pcmod == 0x2)
-   ps_desired = POWER_STATE_3;
-   break;
-   case NVME_POWER_STATE_STANDBY:
-   /* Action unspecified if POWER CONDITION MODIFIER != [0,1] */
-   if (pcmod == 0x0)
-   ps_desired = max(POWER_STATE_0, (lowest_pow_st - 2));
-   else if (pcmod == 0x1)
-   ps_desired = max(POWER_STATE_0, (lowest_pow_st - 1));
-   break;
-   case NVME_POWER_STATE_LU_CONTROL:
-   default:
-   res = nvme_trans_completion(hdr, SAM_STAT_CHECK_CONDITION,
-   ILLEGAL_REQUEST, SCSI_ASC_INVALID_CDB,
-   SCSI_ASCQ_CAUSE_NOT_REPORTABLE);
-   break;
-   }
-   nvme_sc = nvme_set_features(ns->ctrl, NVME_FEAT_POWER_MGMT, ps_desired, 
0,
-   NULL);
-   return nvme_trans_status_code(hdr, nvme_sc);
-}
-
 static int nvme_trans_send_activate_fw_cmd(struct nvme_ns *ns, struct 
sg_io_hdr *hdr,
u8 buffer_id)
 {
@@ -2235,11 +2168,10 @@ static int nvme_trans_synchronize_cache(struct nvme_ns 
*ns,
 static int nvme_trans_start_stop(struct nvme_ns *ns, struct sg_io_hdr *hdr,
u8 *cmd)
 {
-   u8 immed, pcmod, pc, no_flush, start;
+   u8 immed, pcmod, no_flush, start;
 
immed = cmd[1] & 0x01;
pcmod = cmd[3] & 0x0f;
-   pc = (cmd[4] & 0xf0) >> 4;
no_flush = cmd[4]

Re: [PATCH v3 1/3] nvme/scsi: Remove power management support

2016-09-19 Thread Johannes Thumshirn
On Fri, Sep 16, 2016 at 04:25:00PM +0200, Christoph Hellwig wrote:
> On Fri, Sep 16, 2016 at 10:33:02AM -0400, Keith Busch wrote:
> > On Fri, Sep 16, 2016 at 10:33:33AM +0200, Christoph Hellwig wrote:
> > > On Thu, Sep 15, 2016 at 10:24:19PM -0700, Andy Lutomirski wrote:
> > > > Since this code appears to be useless, just delete it.
> > > 
> > > Yes, please!
> > 
> > For a future patch, any objection to deleting the whole file? :)
> 
> Not from me for sure, but I fear some users might be using it after all.

IIRC there was some issues with openSUSE and pre NVMe 1.2 Adapters
without the scsi translation.

Byte,
Johannes
-- 
Johannes Thumshirn  Storage
jthumsh...@suse.de+49 911 74053 689
SUSE LINUX GmbH, Maxfeldstr. 5, 90409 Nürnberg
GF: Felix Imendörffer, Jane Smithard, Graham Norton
HRB 21284 (AG Nürnberg)
Key fingerprint = EC38 9CAB C2C4 F25D 8600 D0D0 0393 969D 2D76 0850


Re: [PATCH v3 1/3] nvme/scsi: Remove power management support

2016-09-16 Thread Christoph Hellwig
On Thu, Sep 15, 2016 at 10:24:19PM -0700, Andy Lutomirski wrote:
> As far as I can tell, there is basically nothing correct about this
> code.  It misinterprets npss (off-by-one).  It hardcodes a bunch of
> power states, which is nonsense, because they're all just indices
> into a table that software needs to parse.  It completely ignores
> the distinction between operational and non-operational states.
> And, until 4.8, if all of the above magically succeeded, it would
> dereference a NULL pointer and OOPS.
> 
> Since this code appears to be useless, just delete it.

Yes, please!

Acked-by: Christoph Hellwig 


Re: [PATCH v3 1/3] nvme/scsi: Remove power management support

2016-09-16 Thread Keith Busch
On Fri, Sep 16, 2016 at 10:33:33AM +0200, Christoph Hellwig wrote:
> On Thu, Sep 15, 2016 at 10:24:19PM -0700, Andy Lutomirski wrote:
> > Since this code appears to be useless, just delete it.
> 
> Yes, please!

For a future patch, any objection to deleting the whole file? :)


Re: [PATCH v3 1/3] nvme/scsi: Remove power management support

2016-09-16 Thread Christoph Hellwig
On Fri, Sep 16, 2016 at 10:33:02AM -0400, Keith Busch wrote:
> On Fri, Sep 16, 2016 at 10:33:33AM +0200, Christoph Hellwig wrote:
> > On Thu, Sep 15, 2016 at 10:24:19PM -0700, Andy Lutomirski wrote:
> > > Since this code appears to be useless, just delete it.
> > 
> > Yes, please!
> 
> For a future patch, any objection to deleting the whole file? :)

Not from me for sure, but I fear some users might be using it after all.



Re: [PATCH v3 1/3] nvme/scsi: Remove power management support

2016-09-16 Thread J Freyensee
On Fri, 2016-09-16 at 10:33 +0200, Christoph Hellwig wrote:
> On Thu, Sep 15, 2016 at 10:24:19PM -0700, Andy Lutomirski wrote:
> > 
> > As far as I can tell, there is basically nothing correct about this
> > code.  It misinterprets npss (off-by-one).  It hardcodes a bunch of
> > power states, which is nonsense, because they're all just indices
> > into a table that software needs to parse.  It completely ignores
> > the distinction between operational and non-operational states.
> > And, until 4.8, if all of the above magically succeeded, it would
> > dereference a NULL pointer and OOPS.
> > 
> > Since this code appears to be useless, just delete it.
> 
> Yes, please!
> 
> Acked-by: Christoph Hellwig 

Acked-by: Jay Freyensee