Commit graph

932,869 commits

Author SHA1 Message Date
Michael Straube
fa1deb227d staging: rtl8188eu: remove unused IS_MCAST
Remove the now unused IS_MCAST.

Signed-off-by: Michael Straube <straube.linux@gmail.com>
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:20:48 +02:00
Michael Straube
2bd827a8d8 staging: rtl8188eu: use is_multicast_ether_addr in rtw_xmit.c
Use is_multicast_ether_addr instead of custom IS_MCAST in
core/rtw_xmit.c. In all uses the address array is properly
aligned.

Signed-off-by: Michael Straube <straube.linux@gmail.com>
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:20:48 +02:00
Michael Straube
e232f8841e staging: rtl8188eu: use is_multicast_ether_addr in rtw_recv.c
Use is_multicast_ether_addr instead of custom IS_MCAST in
core/rtw_recv.c. In all uses the address array/memory is
properly aligned.

Signed-off-by: Michael Straube <straube.linux@gmail.com>
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:20:48 +02:00
Michael Straube
3bd589eb93 staging: rtl8188eu: use is_multicast_ether_addr in rtw_security.c
Use is_multicast_ether_addr instead of custom IS_MCAST in
core/rtw_security.c. In all uses the address array is properly
aligned.

Signed-off-by: Michael Straube <straube.linux@gmail.com>
Reviewed-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:20:47 +02:00
John Whitmore
f1d1871939 staging:rtl8192u: Rename ToLegalChannel - Style
Rename the function ToLegalChannel, which causes a checkpatch issue due
to its use of CamelCase naming. The function has been renamed to
to_legal_channel.

This is a coding style change which should have no impact on runtime
code execution.

Signed-off-by: John Whitmore <johnfwhitmore@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:17:08 +02:00
John Whitmore
b56b1438a1 staging:rtl8192u: Rename IsLegalChannel - Style
The function IsLegalChannel causes a checkpatch issue due to its use
of CamelCase naming. The function has been renamed to is_legal_channel.

This is a coding style change which should have no impact on runtime
code execution.

Signed-off-by: John Whitmore <johnfwhitmore@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:17:08 +02:00
John Whitmore
e859737a15 staging:rtl8192u: Rename DOT11D_ScanComplete - Style
The function DOT11D_ScanComplete causes a checkpatch issue due to its
use of CamelCase naming. The function has been renamed to
dot11d_scan_complete.

This is a coding style change which should have no impact on runtime
code execution.

Signed-off-by: John Whitmore <johnfwhitmore@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:17:08 +02:00
John Whitmore
715a3d9336 staging:rtl8192u: Rename Channel - Style
The function dot11d_get_max_tx_pwr_in_dbm() uses a parameter name,
Channel, which causes a checkpatch issue with CamelCase naming. The
parameter has been renamed to channel.

The change is a coding style change which should have no impact on
runtime code execution.

Signed-off-by: John Whitmore <johnfwhitmore@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:17:08 +02:00
John Whitmore
869b437267 staging:rtl8192u: Rename DOT11D_GetMaxTxPwrInDbm - Style
The function DOT11D_GetMaxTxPwrInDbm causes a checkpatch issue due to
its use of CamelCase naming. The function has been renamed to
dot11d_get_max_tx_pwr_in_dbm.

This is a coding style change which should have no impact on runtime
code execution.

Signed-off-by: John Whitmore <johnfwhitmore@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:17:08 +02:00
John Whitmore
b83ff82670 staging:rtl8192u: Rename dot11d_update_country_ie() parameters - Style
The function dot11d_update_country_ie() uses parameter names which
cause a checkpatch issue due to the use of CamelCase naming.

This is a coding style change which should have no impact on runtime
code execution.

Signed-off-by: John Whitmore <johnfwhitmore@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:17:08 +02:00
John Whitmore
d1f91e0439 staging:rtl8192u: Rename Dot11d_UpdateCountryIe - Style
The function Dot11d_UpdateCountryIe causes a checkpatch issue due to
its use of CamelCase naming, the function has been renamed to
dot11d_update_country_ie.

This is a coding style change which should have no impact on runtime
code execution.

Signed-off-by: John Whitmore <johnfwhitmore@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:17:08 +02:00
John Whitmore
6a184731c2 staging:rtl8192u: Rename Dot11d_Reset - Style
The function Dot11d_Reset causes a checkpatch issue due to its use of
CamelCase naming, so has been renamed to dot11d_reset.

This is a coding style change which should have no impact on runtime
code execution.

Signed-off-by: John Whitmore <johnfwhitmore@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:17:08 +02:00
John Whitmore
c01f06b750 staging:rtl8192u: Rename Dot11d_Init - Style
The function Dot11d_Init() causes a checkpatch issue due to its use
of CamelCase naming. The function has been renamed to dot11d_init.

This is a coding style change which should have no impact on runtime
code execution.

Signed-off-by: John Whitmore <johnfwhitmore@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:17:08 +02:00
John Whitmore
bbb1511665 staging:rtl8192u: Lines should not end with a '(' - Style
A number of function prototypes cause a checkpatch issue - "Lines
should not end with a '(' ". This issue has been cleared by moving
function prototype parameters to the same line as the function name.

This is a coding style change which should have no impact on runtime
code execution.

Signed-off-by: John Whitmore <johnfwhitmore@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:17:08 +02:00
John Whitmore
bc3ae2415b staging:rtl8192u: Rename __pTa - Style
The macro parameter __pTa causes a checkpatch issue because of its use
of CamelCase naming. The parameter has this been renamed to addr.

This is a coding style change which should have no impact on runtime
code execution.

Signed-off-by: John Whitmore <johnfwhitmore@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:17:07 +02:00
John Whitmore
abdc4cae86 staging:rtl8192u: Rename macro parameter __pIeeeDev - Style
Multiple macros, defined in dot11d.h, use the variable __pIeeeDev as a
parameter. This name causes a checkpatch issue due to its use of
CamelCase naming. The parameter name has been changed to ieee_dev to
clear this issue.

This is a coding style change which should have no impact on runtime
code execution.

Signed-off-by: John Whitmore <johnfwhitmore@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:17:07 +02:00
John Whitmore
51e1eb7427 staging:rtl8192u: Replace magic number 6 with ETH_ALEN - Style
The array size 6 represents the length of an Ethernet address so the
magic number has been replaced with the defined constant representing
that length.

This is a coding style change which should not impact runtime code
execution.

Signed-off-by: John Whitmore <johnfwhitmore@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:17:07 +02:00
John Whitmore
ace3c2a23c staging:rtl8192u: Remove macro cpMacAddr - Style
The macro cpMacAddr implements the same functionality as the
ether_addr_copy function defined in etherdevice.h, as a result the
macro has been removed from the code, and its use replaced with the
function call.

Note that the header comment of the function ether_addr_copy states
that the two byte arrays passed as parameters to the function must
be u16 aligned. Because of this the structure ieee80211_network has
had its member variables re-ordered so that the members bssid and
MBssid are aligned to a u16 boundary.

This is a coding style change which should have no impact on runtime
code execution.

Suggested-by: Joe Perches <joe@perches.com>
Signed-off-by: John Whitmore <johnfwhitmore@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:17:07 +02:00
John Whitmore
9391bea9ad staging:rtl8192u: Remove macro eqMacAddr - Style
The macro eqMacAddr implements the same functionality as the
ether_addr_equal function defined in etherdevice.h, as a result the
macro has been removed from the code, and its use replaced with the
function call.

Note that to use the ether_addr_equal function the data array
containing the MAC Address has to be u16 aligned. Because of this the
struct rt_dot11d_info has had it's member variables re-ordered.

This is a coding style change which should have no impact on runtime
code execution.

Suggested-by: Joe Perches <joe@perches.com>
Signed-off-by: John Whitmore <johnfwhitmore@gmail.com>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:17:07 +02:00
Dan Carpenter
b9eb3b14f1 scsi: aacraid: fix a signedness bug
The problem is that ->reset_state is a u8 but it can be set to -1 or -2 in
aac_tmf_callback() and the error handling in aac_eh_target_reset() relies
on it to be signed.

[mkp: fixed typo]

Fixes: 0d643ff3c3 ("scsi: aacraid: use aac_tmf_callback for reset fib")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2018-08-27 13:17:00 -04:00
Ming Lei
d772a65d8a Revert "scsi: core: avoid host-wide host_busy counter for scsi_mq"
This reverts commit 328728630d.

There is fundamental issue in commit 328728630d (scsi: core: avoid
host-wide host_busy counter for scsi_mq) because SCSI's host busy counter
may not be same with counter of blk-mq's inflight tags, especially in case
of none io scheduler.

We may switch to other approach for addressing this scsi_mq's performance
issue, such as percpu counter or kind of ways, so revert this commit first
for fixing this kind of issue in EH path, as reported by Jens.

Cc: Omar Sandoval <osandov@fb.com>,
Cc: "Martin K. Petersen" <martin.petersen@oracle.com>,
Cc: James Bottomley <james.bottomley@hansenpartnership.com>,
Cc: Christoph Hellwig <hch@lst.de>,
Cc: Don Brace <don.brace@microsemi.com>
Cc: Kashyap Desai <kashyap.desai@broadcom.com>
Cc: Mike Snitzer <snitzer@redhat.com>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Laurence Oberman <loberman@redhat.com>
Cc: Bart Van Assche <bart.vanassche@wdc.com>
Cc: Jens Axboe <axboe@kernel.dk>
Reported-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2018-08-27 13:17:00 -04:00
Arnd Bergmann
f45b8934b9 staging: wilc1000: revert "fix TODO to compile spi and sdio components in single module"
The TODO item named "make spi and sdio components coexist in one build"
was apparently addressed a long time ago, but never removed from the
TODO file. However, the new patch that tries to address it actually
makes it worse again by duplicating the common parts of the driver into
two separate modules rather than sharing them. This also introduces a
build regression when one of the two is built-in while the other is a
loadable module:

drivers/staging/wilc1000/wilc_debugfs.o:(.data+0x10): undefined reference to `__this_module'

Reverting the patch makes it build again. I'm leaving the TODO file
modification though, as there is nothing left to do for this item.

A related problem however still seems to exist: one still cannot have
multiple concurrent instances of wilc1000 devices present in the
system, as there are lots of shared global variables such as

host_interface.c:static struct wilc_vif *periodic_rssi_vif;
wilc_sdio.c:static struct wilc_sdio g_sdio;
wilc_wlan.c:static enum chip_ps_states chip_ps_state = CHIP_WAKEDUP;
wilc_wlan.c:static u32 pending_acks;
wilc_wfi_cfgoperations.c:int wilc_connecting;

In order to have multiple instances working (sdio, spi, or mixed),
all such variables need to be dynamically allocated per instance and
stored in 'struct wilc' or one of the structures referenced by it.

Fixes: 9abc44ba4e ("staging: wilc1000: fix TODO to compile spi and sdio components in single module")
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
2018-08-27 19:12:59 +02:00
Ming Lei
23aa8e69f2 Revert "scsi: core: fix scsi_host_queue_ready"
This reverts commit 265d59aacb.

There is fundamental issue in commit 328728630d (scsi: core: avoid
host-wide host_busy counter for scsi_mq) because SCSI's host busy counter
may not be same with counter of blk-mq's inflight tags, especially in case
of none io scheduler.

So revert this commit first.

Cc: Omar Sandoval <osandov@fb.com>,
Cc: "Martin K. Petersen" <martin.petersen@oracle.com>,
Cc: James Bottomley <james.bottomley@hansenpartnership.com>,
Cc: Christoph Hellwig <hch@lst.de>,
Cc: Don Brace <don.brace@microsemi.com>
Cc: Kashyap Desai <kashyap.desai@broadcom.com>
Cc: Mike Snitzer <snitzer@redhat.com>
Cc: Hannes Reinecke <hare@suse.de>
Cc: Laurence Oberman <loberman@redhat.com>
Cc: Bart Van Assche <bart.vanassche@wdc.com>
Cc: Guenter Roeck <linux@roeck-us.net>
Cc: Jens Axboe <axboe@kernel.dk>
Reported-by: Jens Axboe <axboe@kernel.dk>
Signed-off-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2018-08-27 12:59:43 -04:00
Geert Uytterhoeven
4e8065aa6c scsi: libata: Add missing newline at end of file
With gcc 4.1.2:

    drivers/ata/libata-core.c:7396:33: warning: no newline at end of file

Fixes: 2fa4a32613 ("scsi: libsas: dynamically allocate and free ata host")
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2018-08-27 12:26:10 -04:00
Varun Prakash
cedefa8544 scsi: target: iscsi: cxgbit: use pr_debug() instead of pr_info()
DDP programming happens in data path and it can fail because of lack of
resources so use pr_debug() instead of pr_info() for this case.

Signed-off-by: Varun Prakash <varun@chelsio.com>
Reviewed-by: Mike Christie <mchristi@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2018-08-27 12:26:10 -04:00
Martin Wilck
eb53a3ea3e scsi: hpsa: limit transfer length to 1MB, not 512kB
e2c7b43 was supposed to limit transfer length to 1MB, but got the unit of
max_sectors wrong.

Fixes: e2c7b433f7 ("scsi: hpsa: limit transfer length to 1MB")
Signed-off-by: Martin Wilck <mwilck@suse.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2018-08-27 12:26:10 -04:00
James Smart
53e13ee087 scsi: lpfc: Correct MDS diag and nvmet configuration
A recent change added some MDS processing in the lpfc_drain_txq routine
that relies on the fcp_wq being allocated. For nvmet operation the fcp_wq
is not allocated because it can only be an nvme-target.  When the original
MDS support was added LS_MDS_LOOPBACK was defined wrong, (0x16) it should
have been 0x10 (decimal value used for hex setting). This incorrect value
allowed MDS_LOOPBACK to be set simultaneously with LS_NPIV_FAB_SUPPORTED,
causing the driver to crash when it accesses the non-existent fcp_wq.

Correct the bad value setting for LS_MDS_LOOPBACK.

Fixes: 	ae9e28f36a  ("lpfc: Add MDS Diagnostic support.")
Cc: <stable@vger.kernel.org> # v4.12+
Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
Tested-by: Ewan D. Milne <emilne@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2018-08-27 12:26:10 -04:00
James Smart
9abd9990e9 scsi: lpfc: Default fdmi_on to on
Change default behavior for fdmi registration to on.

[mkp: patch was mangled]

Signed-off-by: Dick Kennedy <dick.kennedy@broadcom.com>
Signed-off-by: James Smart <james.smart@broadcom.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2018-08-27 12:26:10 -04:00
Varun Prakash
68bdc63072 scsi: csiostor: fix incorrect port capabilities
- use be32_to_cpu() instead of ntohs() for 32 bit port capabilities.

 - add a new function fwcaps32_to_caps16() to convert 32 bit port
   capabilities to 16 bit port capabilities.

Signed-off-by: Varun Prakash <varun@chelsio.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2018-08-27 12:26:10 -04:00
Varun Prakash
89809b028b scsi: csiostor: add a check for NULL pointer after kmalloc()
Reported-by: Colin Ian King <colin.king@canonical.com>
Signed-off-by: Varun Prakash <varun@chelsio.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2018-08-27 12:26:10 -04:00
John Pittman
a7ccd92c8d scsi: documentation: add scsi_mod.use_blk_mq to scsi-parameters
Kernel line argument scsi_mod.use_blk_mq is missing from file
Documentation/scsi/scsi-parameters.txt.  Add this option, providing
mention of config setting and format.

[mkp: clarified where to look]

Signed-off-by: John Pittman <jpittman@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2018-08-27 12:26:10 -04:00
Geert Uytterhoeven
3fba68fa35 scsi: core: Update SCSI_MQ_DEFAULT help text to match default
The default was changed, but the help text was not updated.

Fix grammar (s/the option/this option/) while at it.

[mkp: drop "new" as suggested by John Garry]

Fixes: d5038a13ec ("scsi: core: switch to scsi-mq by default")
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be>
Reviewed-by: Ming Lei <ming.lei@redhat.com>
Signed-off-by: Martin K. Petersen <martin.petersen@oracle.com>
2018-08-27 12:26:10 -04:00
Juergen Gross
6f2f39ad1a xen/blkback: remove unused pers_gnts_lock from struct xen_blkif_ring
pers_gnts_lock isn't being used anywhere. Remove it.

Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
2018-08-27 12:12:04 -04:00
Juergen Gross
d77ff24e7f xen/blkback: move persistent grants flags to bool
The struct persistent_gnt flags member is meant to be a bitfield of
different flags. There is only PERSISTENT_GNT_ACTIVE flag left, so
convert it to a bool named "active".

Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
2018-08-27 12:12:04 -04:00
Juergen Gross
4bcddbae01 xen/blkfront: reorder tests in xlblk_init()
In case we don't want pv block devices we should not test parameters
for sanity and eventually print out error messages. So test precluding
conditions before checking parameters.

Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
2018-08-27 12:12:03 -04:00
Juergen Gross
a46b53672b xen/blkfront: cleanup stale persistent grants
Add a periodic cleanup function to remove old persistent grants which
are no longer in use on the backend side. This avoids starvation in
case there are lots of persistent grants for a device which no longer
is involved in I/O business.

Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
2018-08-27 12:12:03 -04:00
Juergen Gross
973e5405f2 xen/blkback: don't keep persistent grants too long
Persistent grants are allocated until a threshold per ring is being
reached. Those grants won't be freed until the ring is being destroyed
meaning there will be resources kept busy which might no longer be
used.

Instead of freeing only persistent grants until the threshold is
reached add a timestamp and remove all persistent grants not having
been in use for a minute.

Signed-off-by: Juergen Gross <jgross@suse.com>
Reviewed-by: Roger Pau Monné <roger.pau@citrix.com>
Signed-off-by: Konrad Rzeszutek Wilk <konrad.wilk@oracle.com>
2018-08-27 12:12:03 -04:00
Huang Rui
07e6d3f03b drm/amdgpu: move PD/PT bos on LRU again
The new bulk moving functionality is ready, the overhead of moving PD/PT bos to
LRU is fixed. So move them on LRU again.

Signed-off-by: Huang Rui <ray.huang@amd.com>
Tested-by: Mike Lothian <mike@fireburn.co.uk>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
Acked-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-08-27 11:11:22 -05:00
Huang Rui
f921661bd4 drm/amdgpu: use bulk moves for efficient VM LRU handling (v6)
I continue to work for bulk moving that based on the proposal by Christian.

Background:
amdgpu driver will move all PD/PT and PerVM BOs into idle list. Then move all of
them on the end of LRU list one by one. Thus, that cause so many BOs moved to
the end of the LRU, and impact performance seriously.

Then Christian provided a workaround to not move PD/PT BOs on LRU with below
patch:
Commit 0bbf32026cf5ba41e9922b30e26e1bed1ecd38ae ("drm/amdgpu: band aid
validating VM PTs")

However, the final solution should bulk move all PD/PT and PerVM BOs on the LRU
instead of one by one.

Whenever amdgpu_vm_validate_pt_bos() is called and we have BOs which need to be
validated we move all BOs together to the end of the LRU without dropping the
lock for the LRU.

While doing so we note the beginning and end of this block in the LRU list.

Now when amdgpu_vm_validate_pt_bos() is called and we don't have anything to do,
we don't move every BO one by one, but instead cut the LRU list into pieces so
that we bulk move everything to the end in just one operation.

Test data:
+--------------+-----------------+-----------+---------------------------------------+
|              |The Talos        |Clpeak(OCL)|BusSpeedReadback(OCL)                  |
|              |Principle(Vulkan)|           |                                       |
+------------------------------------------------------------------------------------+
|              |                 |           |0.319 ms(1k) 0.314 ms(2K) 0.308 ms(4K) |
| Original     |  147.7 FPS      |  76.86 us |0.307 ms(8K) 0.310 ms(16K)             |
+------------------------------------------------------------------------------------+
| Orignial + WA|                 |           |0.254 ms(1K) 0.241 ms(2K)              |
|(don't move   |  162.1 FPS      |  42.15 us |0.230 ms(4K) 0.223 ms(8K) 0.204 ms(16K)|
|PT BOs on LRU)|                 |           |                                       |
+------------------------------------------------------------------------------------+
| Bulk move    |  163.1 FPS      |  40.52 us |0.244 ms(1K) 0.252 ms(2K) 0.213 ms(4K) |
|              |                 |           |0.214 ms(8K) 0.225 ms(16K)             |
+--------------+-----------------+-----------+---------------------------------------+

After test them with above three benchmarks include vulkan and opencl. We can
see the visible improvement than original, and even better than original with
workaround.

v2: move all BOs include idle, relocated, and moved list to the end of LRU and
put them together.
v3: remove unused parameter and use list_for_each_entry instead of the one with
save entry.
v4: move the amdgpu_vm_move_to_lru_tail after command submission, at that time,
all bo will be back on idle list.
v5: remove amdgpu_vm_move_to_lru_tail_by_list(), use bulk_moveable instread of
validated, and move ttm_bo_bulk_move_lru_tail() also into
amdgpu_vm_move_to_lru_tail().
v6: clean up and fix return value.

Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Huang Rui <ray.huang@amd.com>
Tested-by: Mike Lothian <mike@fireburn.co.uk>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
Acked-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-08-27 11:11:22 -05:00
Huang Rui
7748e2dcda drm/ttm: add bulk move function on LRU
This function allow us to bulk move a group of BOs to the tail of their LRU.
The positions of group of BOs are stored on the (first, last) bulk_move_pos
structure.

Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Huang Rui <ray.huang@amd.com>
Tested-by: Mike Lothian <mike@fireburn.co.uk>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
Acked-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-08-27 11:11:21 -05:00
Christian König
9a2779528e drm/ttm: revise ttm_bo_move_to_lru_tail to support bulk moves
When move a BO to the end of LRU, it need remember the BO positions.
Make sure all moved bo in between "first" and "last". And they will be bulk
moving together.

Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Huang Rui <ray.huang@amd.com>
Tested-by: Mike Lothian <mike@fireburn.co.uk>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
Acked-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-08-27 11:11:21 -05:00
Christian König
8c7655a0fd drm/ttm: add helper structures for bulk moves on lru list
Add bulk move pos to store the pointer of first and last buffer object.
The list in between will be bulk moved on lru list.

Signed-off-by: Christian König <christian.koenig@amd.com>
Signed-off-by: Huang Rui <ray.huang@amd.com>
Tested-by: Mike Lothian <mike@fireburn.co.uk>
Tested-by: Dieter Nützel <Dieter@nuetzel-hh.de>
Acked-by: Chunming Zhou <david1.zhou@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-08-27 11:11:20 -05:00
Christian König
11c3a249ff drm/amdgpu: add amdgpu_gmc_pd_addr helper
Add a helper to get the root PD address and remove the workarounds from
the GMC9 code for that.

Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Felix Kuehling <Felix.Kuehling@amd.com>
Reviewed-by: Alex Deucher <alexander.deucher@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-08-27 11:11:19 -05:00
Christian König
4e830fb1b5 drm/amdgpu: remove gart.table_addr
We can easily figure out the address on the fly.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-08-27 11:11:19 -05:00
Christian König
1123b989c5 drm/amdgpu: rename gart.robj into gart.bo
sed -i "s/gart.robj/gart.bo/" drivers/gpu/drm/amd/amdgpu/*.c
sed -i "s/gart.robj/gart.bo/" drivers/gpu/drm/amd/amdgpu/*.h

Just cleaning up radeon leftovers.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-08-27 11:11:18 -05:00
Christian König
cbd5285142 drm/amdgpu: move setting the GART addr into TTM
Move setting the GART addr for window based copies into the TTM code who
uses it.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-08-27 11:11:18 -05:00
Christian König
9a02ece43e drm/amdgpu: cleanup VM handling in the CS a bit
Add a helper function for getting the root PD addr and cleanup join the
two VM related functions and cleanup the function name.

No functional change.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-08-27 11:11:17 -05:00
Christian König
262b9c392e drm/amdgpu: validate the VM root PD from the VM code
Preparation for following changes. This validates the root PD twice,
but the overhead of that should be minimal.

Signed-off-by: Christian König <christian.koenig@amd.com>
Reviewed-by: Junwei Zhang <Jerry.Zhang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-08-27 11:11:17 -05:00
Christian König
12938fad23 drm/amdgpu: cleanup GPU recovery check a bit (v2)
Check if we should call the function instead of providing the forced
flag.

v2: rebase on KFD changes (Alex)

Signed-off-by: Christian König <christian.koenig@amd.com>
Acked-by: Andrey Grodzovsky <andrey.grodzovsky@amd.com>
Reviewed-by: Huang Rui <ray.huang@amd.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-08-27 11:11:16 -05:00
kbuild test robot
1849e73748 drm/amdgpu: amdgpu_kiq_reg_write_reg_wait() can be static
Fixes: d790449835e6 ("drm/amdgpu: use kiq to do invalidate tlb")
Reviewed-by: Emily Deng <Emily.Deng@amd.com>
Signed-off-by: kbuild test robot <fengguang.wu@intel.com>
Signed-off-by: Alex Deucher <alexander.deucher@amd.com>
2018-08-27 11:11:16 -05:00