]> www.infradead.org Git - users/jedix/linux-maple.git/log
users/jedix/linux-maple.git
8 years agox86/platform/UV: Add obtaining GAM Range Table from UV BIOS
Mike Travis [Fri, 29 Apr 2016 21:54:18 +0000 (16:54 -0500)]
x86/platform/UV: Add obtaining GAM Range Table from UV BIOS

Orabug: 25477822

UV4 uses a GAM (globally addressed memory) architecture that supports
variable sized memory per node.  This replaces the old "M" value (number
of address bits per node) with a range table for conversions between
addresses and physical node (pnode) id's.  This table is obtained from UV
BIOS via the EFI UVsystab table.  Support for older EFI UVsystab tables
is maintained.

Tested-by: Dimitri Sivanich <sivanich@sgi.com>
Tested-by: John Estabrook <estabrook@sgi.com>
Tested-by: Gary Kroening <gfk@sgi.com>
Tested-by: Nathan Zimmer <nzimmer@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Cc: Andrew Banman <abanman@sgi.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Len Brown <len.brown@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russ Anderson <rja@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20160429215405.329827545@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit ef93bf803999445985acb25f4ed8772e1aa81221)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
8 years agox86/platform/UV: Add UV4 addressing discovery function
Mike Travis [Fri, 29 Apr 2016 21:54:17 +0000 (16:54 -0500)]
x86/platform/UV: Add UV4 addressing discovery function

Orabug: 25477822

UV4 requires early system wide addressing values.  This involves the use
of the CPUID instruction to obtain these values.  The current function
(detect_extended_topology()) in the kernel has been copied and streamlined,
with the limitation that only CPU's used by UV architectures are supported.

Tested-by: John Estabrook <estabrook@sgi.com>
Tested-by: Gary Kroening <gfk@sgi.com>
Tested-by: Nathan Zimmer <nzimmer@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Dimitri Sivanich <sivanich@sgi.com>
Cc: Andrew Banman <abanman@sgi.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Len Brown <len.brown@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russ Anderson <rja@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20160429215405.155660884@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit 405422d88c686e88b4241c0201fd96b61ab8bd77)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
8 years agox86/platform/UV: Fold blade info into per node hub info
Mike Travis [Fri, 17 Feb 2017 17:08:37 +0000 (12:08 -0500)]
x86/platform/UV: Fold blade info into per node hub info
 structs

Orabug: 25477822

Migrate references from the blade info structs to the per node hub info
structs.  This phases out the allocation of the list of per blade info
structs on node 0, in favor of a per node hub info struct allocated on
the node's local memory.

There are also some minor cosemetic changes in the comments and whitespace
to clean things up a bit.

Tested-by: Dimitri Sivanich <sivanich@sgi.com>
Tested-by: John Estabrook <estabrook@sgi.com>
Tested-by: Gary Kroening <gfk@sgi.com>
Tested-by: Nathan Zimmer <nzimmer@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Andrew Banman <abanman@sgi.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Len Brown <len.brown@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russ Anderson <rja@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20160429215404.987204515@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit 906f3b20da8c6ec3eeef81753b4af9b6780e2edc)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
8 years agox86/platform/UV: Allocate common per node hub info structs on local node
Mike Travis [Fri, 29 Apr 2016 21:54:15 +0000 (16:54 -0500)]
x86/platform/UV: Allocate common per node hub info structs on local node

Orabug: 25477822

Allocate and setup per node hub info structs.  CPU 0/Node 0 hub info
is statically allocated to be accessible early in system startup.  The
remaining hub info structs are allocated on the node's local memory,
and shared among the CPU's on that node.  This leaves the small amount
of info unique to each CPU in the per CPU info struct.

Memory is saved by combining the common per node info fields to common
node local structs.  In addtion, since the info is read only only after
setup, it should stay in the L3 cache of the local processor socket.
This should therefore improve the cache hit rate when a group of cpus
on a node are all interrupted for a common task.

Tested-by: John Estabrook <estabrook@sgi.com>
Tested-by: Gary Kroening <gfk@sgi.com>
Tested-by: Nathan Zimmer <nzimmer@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Dimitri Sivanich <sivanich@sgi.com>
Reviewed-by: Andrew Banman <abanman@sgi.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Len Brown <len.brown@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russ Anderson <rja@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20160429215404.813051625@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit 3edcf2ff7ae50d1096030fab9a1bafb421e07d4c)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
8 years agox86/platform/UV: Move blade local processor ID to the per cpu info struct
Mike Travis [Fri, 29 Apr 2016 21:54:14 +0000 (16:54 -0500)]
x86/platform/UV: Move blade local processor ID to the per cpu info struct

Orabug: 25477822

Move references to blade local processor ID to the new per cpu info
structs.  Create an access function that makes this move, and other
potential moves opaque to callers of this function.  Define a flag
that indicates to callers in external GPL modules that this function
replaces any local definition.  This allows calling source code to be
built for both pre-UV4 kernels as well as post-UV4 kernels.

Tested-by: John Estabrook <estabrook@sgi.com>
Tested-by: Gary Kroening <gfk@sgi.com>
Tested-by: Nathan Zimmer <nzimmer@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Dimitri Sivanich <sivanich@sgi.com>
Cc: Andrew Banman <abanman@sgi.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Len Brown <len.brown@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russ Anderson <rja@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20160429215404.644173122@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit 5627a8251f7861175b193a44dc3d8cb478d1135a)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
8 years agox86/platform/UV: Move scir info to the per cpu info struct
Mike Travis [Fri, 29 Apr 2016 21:54:13 +0000 (16:54 -0500)]
x86/platform/UV: Move scir info to the per cpu info struct

Orabug: 25477822

Change the references to the SCIR fields to the new per cpu info structs.

Tested-by: John Estabrook <estabrook@sgi.com>
Tested-by: Gary Kroening <gfk@sgi.com>
Tested-by: Nathan Zimmer <nzimmer@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Dimitri Sivanich <sivanich@sgi.com>
Cc: Andrew Banman <abanman@sgi.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Len Brown <len.brown@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russ Anderson <rja@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20160429215404.452538234@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit d38bb135d814e96811e1a0778564d7a2df922e28)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
8 years agox86/platform/UV: Create per cpu info structs to replace per hub info structs
Mike Travis [Fri, 29 Apr 2016 21:54:12 +0000 (16:54 -0500)]
x86/platform/UV: Create per cpu info structs to replace per hub info structs

Orabug: 25477822

The major portion of the hub info is common to all cpus on that hub.
This is step one of moving the per cpu hub info to a per node hub info
struct.  This patch creates the small per cpu info struct that will
contain only information specific to each CPU.

Tested-by: John Estabrook <estabrook@sgi.com>
Tested-by: Gary Kroening <gfk@sgi.com>
Tested-by: Nathan Zimmer <nzimmer@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Dimitri Sivanich <sivanich@sgi.com>
Cc: Andrew Banman <abanman@sgi.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Len Brown <len.brown@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russ Anderson <rja@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20160429215404.282265563@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit 0045ddd23f21ad1964c01228257bc6c692e1c2f9)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
8 years agox86/platform/UV: Update MMIOH setup function to work for both UV3 and UV4
Mike Travis [Fri, 29 Apr 2016 21:54:11 +0000 (16:54 -0500)]
x86/platform/UV: Update MMIOH setup function to work for both UV3 and UV4

Orabug: 25477822

Since UV3 and UV4 MMIOH regions are setup the same, we can use a common
function to setup both.

Tested-by: John Estabrook <estabrook@sgi.com>
Tested-by: Gary Kroening <gfk@sgi.com>
Tested-by: Nathan Zimmer <nzimmer@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Dimitri Sivanich <sivanich@sgi.com>
Cc: Andrew Banman <abanman@sgi.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Len Brown <len.brown@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russ Anderson <rja@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20160429215404.100504077@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit a2f28e6950cec75320a8c3c8747a6e3ad08cfd2b)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
8 years agox86/platform/UV: Clean up redunduncies after merge of UV4 MMR definitions
Mike Travis [Fri, 29 Apr 2016 21:54:10 +0000 (16:54 -0500)]
x86/platform/UV: Clean up redunduncies after merge of UV4 MMR definitions

Orabug: 25477822

Clean up any redundancies caused by new UV4 MMR definitions superseding
any previously definitions local to functions.

Tested-by: John Estabrook <estabrook@sgi.com>
Tested-by: Gary Kroening <gfk@sgi.com>
Tested-by: Nathan Zimmer <nzimmer@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Dimitri Sivanich <sivanich@sgi.com>
Reviewed-by: Andrew Banman <abanman@sgi.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Len Brown <len.brown@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russ Anderson <rja@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20160429215403.934728974@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit b608f87fe886d3ef7f9f8eb8ba58f45beb61c286)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
8 years agox86/platform/UV: Add UV4 Specific MMR definitions
Mike Travis [Fri, 29 Apr 2016 21:54:08 +0000 (16:54 -0500)]
x86/platform/UV: Add UV4 Specific MMR definitions

Orabug: 25477822

This adds the MMR definitions for UV4 via an automated script that uses
the output from a hardware verilog code to symbol converter.  The large
number of insertions is caused by the UV4 design changing many similarly
named fields in MMR's that are named the same.  This prompted the extra
production of architecture dependent field defines.

Tested-by: John Estabrook <estabrook@sgi.com>
Tested-by: Gary Kroening <gfk@sgi.com>
Tested-by: Nathan Zimmer <nzimmer@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Cc: Andrew Banman <abanman@sgi.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Dimitri Sivanich <sivanich@sgi.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Len Brown <len.brown@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russ Anderson <rja@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20160429215403.580158916@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit 0f0d84c08d38cc4c61fc04f94db0713aa82a39bc)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
8 years agox86/platform/UV: Prep for UV4 MMR updates
Mike Travis [Fri, 17 Feb 2017 17:06:14 +0000 (12:06 -0500)]
x86/platform/UV: Prep for UV4 MMR updates

Orabug: 25477822

Cleanup patch to rearrange code and modify some defines so the next
patch, the new UV4 MMR definitions can be merged cleanly.

* Clean up the M/N related address constants (M is # of address bits per
  blade, N is the # of blade selection bits per SSI/partition).

* Fix the lookup of the alias overlay addresses and NMI definitions to
  allow for flexibility in newer UV architecture types.

Tested-by: John Estabrook <estabrook@sgi.com>
Tested-by: Gary Kroening <gfk@sgi.com>
Tested-by: Nathan Zimmer <nzimmer@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Dimitri Sivanich <sivanich@sgi.com>
Cc: Andrew Banman <abanman@sgi.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Len Brown <len.brown@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russ Anderson <rja@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20160429215403.401604203@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit c443c03dd0d97620022483be6705ff611695a29c)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
8 years agox86/platform/UV: Add UV MMR Illegal Access Function
Mike Travis [Fri, 29 Apr 2016 21:54:06 +0000 (16:54 -0500)]
x86/platform/UV: Add UV MMR Illegal Access Function

Orabug: 25477822

This new function is generated by the UV MMR generation script to
identify MMR registers and fields that are not defined for a specific
UV architecture.  With this switch, the immediate panic can be replaced
with a message and a bad return value allowing either hardware or the
emulator to diagnose the problem.  It allows functions common to some
UV arches to use common defines that might not be fully defined for all
arches, as long as they do not reference them on the unsupported arches.

Tested-by: John Estabrook <estabrook@sgi.com>
Tested-by: Gary Kroening <gfk@sgi.com>
Tested-by: Nathan Zimmer <nzimmer@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Dimitri Sivanich <sivanich@sgi.com>
Cc: Andrew Banman <abanman@sgi.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Len Brown <len.brown@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russ Anderson <rja@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20160429215403.231926687@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit 7563421b13da21dd7a947f658b5299e65ed95cbe)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
8 years agox86/platform/UV: Add UV4 Specific Defines
Mike Travis [Fri, 29 Apr 2016 21:54:05 +0000 (16:54 -0500)]
x86/platform/UV: Add UV4 Specific Defines

Orabug: 25477822

Add UV4 specific defines to determine if current system type is a
UV4 system.

Tested-by: John Estabrook <estabrook@sgi.com>
Tested-by: Gary Kroening <gfk@sgi.com>
Tested-by: Nathan Zimmer <nzimmer@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Dimitri Sivanich <sivanich@sgi.com>
Cc: Andrew Banman <abanman@sgi.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Len Brown <len.brown@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russ Anderson <rja@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20160429215403.072323684@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit a0ec83f316e1d933d8c820d249972574324c2d25)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
8 years agox86/platform/UV: Add UV Architecture Defines
Mike Travis [Fri, 29 Apr 2016 21:54:04 +0000 (16:54 -0500)]
x86/platform/UV: Add UV Architecture Defines

Orabug: 25477822

Add defines to control which UV architectures are supported, and modify the
'if (is_uvX_*)' functions to return constant 0 for those not supported.
This will help optimize code paths when support for specific UV arches
is removed.

Tested-by: John Estabrook <estabrook@sgi.com>
Tested-by: Gary Kroening <gfk@sgi.com>
Tested-by: Nathan Zimmer <nzimmer@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Reviewed-by: Dimitri Sivanich <sivanich@sgi.com>
Cc: Andrew Banman <abanman@sgi.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Len Brown <len.brown@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russ Anderson <rja@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20160429215402.897143440@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit e0ee1c97c3b1cabc3651d7bcf39c1f54d736fd20)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
8 years agox86/platform/UV: Add Initial UV4 definitions
Mike Travis [Fri, 29 Apr 2016 21:54:03 +0000 (16:54 -0500)]
x86/platform/UV: Add Initial UV4 definitions

Orabug: 25477822

Add preliminary UV4 defines.

Tested-by: John Estabrook <estabrook@sgi.com>
Tested-by: Gary Kroening <gfk@sgi.com>
Tested-by: Nathan Zimmer <nzimmer@sgi.com>
Signed-off-by: Mike Travis <travis@sgi.com>
Cc: Andrew Banman <abanman@sgi.com>
Cc: Andrew Morton <akpm@linux-foundation.org>
Cc: Andy Lutomirski <luto@amacapital.net>
Cc: Borislav Petkov <bp@alien8.de>
Cc: Brian Gerst <brgerst@gmail.com>
Cc: Denys Vlasenko <dvlasenk@redhat.com>
Cc: Dimitri Sivanich <sivanich@sgi.com>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Len Brown <len.brown@intel.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russ Anderson <rja@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/20160429215402.703593187@asylum.americas.sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit eb1e3461b8912ce4794fd2b7b414338a59461601)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
8 years agox86/platform/uv: Disable UV BAU by default
Alex Thorlton [Thu, 31 Mar 2016 19:18:29 +0000 (14:18 -0500)]
x86/platform/uv: Disable UV BAU by default

Orabug: 25477822

For several years, the common practice has been to boot UVs with the
"nobau" parameter on the command line, to disable the BAU.  We've
decided that it makes more sense to just disable the BAU by default in
the kernel, and provide the option to turn it on, if desired.

For now, having the on/off switch doesn't buy us any more than just
reversing the logic would, but we're working towards having the BAU
enabled by default on UV4.  When those changes are in place, having the
on/off switch will make more sense than an enable flag, since the
default behavior will be different depending on the system version.

I've also added a bit of documentation for the new parameter to
Documentation/kernel-parameters.txt.

Signed-off-by: Alex Thorlton <athorlton@sgi.com>
Reviewed-by: Hedi Berriche <hedi@sgi.com>
Cc: Jonathan Corbet <corbet@lwn.net>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Link: http://lkml.kernel.org/r/1459451909-121845-1-git-send-email-athorlton@sgi.com
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit 1c532e00a0c649ac6f0703e8c2e095c9c1d30625)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
8 years agox86/platform/uv: Include clocksource.h for clocksource_touch_watchdog()
Ingo Molnar [Fri, 11 Dec 2015 08:01:30 +0000 (09:01 +0100)]
x86/platform/uv: Include clocksource.h for clocksource_touch_watchdog()

Orabug: 25477822

This build failure triggers on 64-bit allmodconfig:

  arch/x86/platform/uv/uv_nmi.c:493:2: error: implicit declaration of function ‘clocksource_touch_watchdog’ [-Werror=implicit-function-declaration]

which is caused by recent changes exposing a missing clocksource.h include
in uv_nmi.c:

  cc1e24fdb064 x86/vdso: Remove pvclock fixmap machinery

this file got clocksource.h indirectly via fixmap.h - that stealth route
of header inclusion is now gone.

Cc: Borislav Petkov <bp@alien8.de>
Cc: H. Peter Anvin <hpa@zytor.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit d51953b0873358d13b189996e6976dfa12a9b59d)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
8 years agox86/platform/uv: Implement simple dump failover if kdump fails
Mike Travis [Sun, 13 Sep 2015 02:51:10 +0000 (21:51 -0500)]
x86/platform/uv: Implement simple dump failover if kdump fails

Orabug: 25477822

The ability to trigger a kdump using the system NMI command
was added by

    commit 12ba6c990fab ("x86/UV: Add kdump to UV NMI handler")
    Author: Mike Travis <travis@sgi.com>
    Date:   Mon Sep 23 16:25:03 2013 -0500

This is useful because when kdump is working the information
gathered is more informative than the original per CPU stack
traces or "dump" option.  However a number of things can go
wrong with kdump and then the stack traces are more useful than
nothing.

The two most common reasons for kdump to not be available are:

  1) if a problem occurs during boot before the kdump service is
     started, or
  2) the kdump daemon failed to start.

In either case the call to crash_kexec() returns unexpectedly.

When this happens uv_nmi_kdump() also sets the
uv_nmi_kexec_failed flag which causes the slave CPU's to also
return to the NMI handler. Upon this unexpected return to the
NMI handler, the NMI handler will revert to the "dump" action
which uses show_regs() to obtain a process trace dump for all
the CPU's.

Other minor changes:
    The "dump" action now generates both the show_regs() stack trace
    and show instruction pointer information.  Whereas the "ips"
    action only shows instruction pointers for non-idle CPU's.  This
    is more like an abbreviated "ps" display.

    Change printk(KERN_DEFAULT...) --> pr_info()

Signed-off-by: Mike Travis <travis@sgi.com>
Signed-off-by: George Beshers <gbeshers@sgi.com>
Cc: Alex Thorlton <athorlton@sgi.com>
Cc: Dimitri Sivanich <sivanich@sgi.com>
Cc: Hedi Berriche <hedi@sgi.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russ Anderson <rja@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit d0a9964e98731c708500a2e712f28f9d39183647)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
Conflicts:
arch/x86/platform/uv/uv_nmi.c

8 years agox86/platform/uv: Insert per_cpu accessor function on uv_hub_nmi
George Beshers [Sun, 13 Sep 2015 02:51:05 +0000 (21:51 -0500)]
x86/platform/uv: Insert per_cpu accessor function on uv_hub_nmi

Orabug: 25477822

UV: NMI: insert this_cpu_read accessor function on uv_hub_nmi.

On SGI UV systems a 'power nmi' command from the CMC causes
all processors to drop into uv_handle_nmi().  With the 4.0
kernel this results in

    BUG: unable to handle kernel paging request

The bug is caused by the current code trying to use the PER_CPU
variable uv_cpu_nmi.hub without an appropriate accessor
function. That oversight occurred in

    commit e16321709c82 ("uv: Replace __get_cpu_var")
    Author: Christoph Lameter <cl@linux.com>
    Date:   Sun Aug 17 12:30:41 2014 -0500

This patch inserts this_cpu_read() in the uv_hub_nmi macro
restoring the intended functionality.

Signed-off-by: George Beshers <gbeshers@sgi.com>
Acked-by: Mike Travis <travis@sgi.com>
Cc: Alex Thorlton <athorlton@sgi.com>
Cc: Dimitri Sivanich <sivanich@sgi.com>
Cc: Hedi Berriche <hedi@sgi.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Russ Anderson <rja@sgi.com>
Cc: Thomas Gleixner <tglx@linutronix.de>
Cc: linux-kernel@vger.kernel.org
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit 7c52198b9c36001d47e11d50f7f1556805c5c3e3)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
8 years agox86/platform/uv: Make SGI UV dependent on CONFIG_PCI
Ingo Molnar [Wed, 6 May 2015 04:23:59 +0000 (06:23 +0200)]
x86/platform/uv: Make SGI UV dependent on CONFIG_PCI

Orabug: 25477822

Recent PCI changes stopped exporting PCI constants if !CONFIG_PCI,
which made the UV build fail:

  arch/x86/kernel/apic/x2apic_uv_x.c:843:16: error: ‘PCI_VGA_STATE_CHANGE_BRIDGE’ undeclared (first use in this function)
  arch/x86/kernel/apic/x2apic_uv_x.c:1023:2: error: implicit declaration of function ‘pci_register_set_vga_state’ [-Werror=implicit-function-declaration]

As it's unlikely that an UV bootup will get far without PCI
enumeration, make the platform Kconfig switch (CONFIG_X86_UV)
depend on CONFIG_PCI=y.

Cc: Robin Holt <holt@sgi.com>
Cc: Dimitri Sivanich <sivanich@sgi.com>
Cc: Russ Anderson <rja@sgi.com>
Cc: Mike Travis <travis@sgi.com>
Cc: Jack Steiner <steiner@sgi.com>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
(cherry picked from commit 1222e564cf4394af0b3c5e8a73330b20862c068b)
Signed-off-by: Brian Maly <brian.maly@oracle.com>
8 years agoIB/IPoIB: Add destination address when re-queue packet
Yuval Shaia [Sat, 11 Feb 2017 22:22:48 +0000 (17:22 -0500)]
IB/IPoIB: Add destination address when re-queue packet

Orabug: 25466606

[ Pulled from upstream discussion, not yet merged ]

When sending packet to destination that was not resolved yet
via path query, the driver keeps the skb and tries to re-send it
again when the path is resolved.

But when re-sending via dev_queue_xmit the kernel doesn't call
to dev_hard_header, so IPoIB needs to keep 20 bytes in the skb
and to put the destination address inside them.

In that way the dev_start_xmit will have the correct destination,
and the driver won't take the destination from the skb->data, while
nothing exists there, which causes to packet be be dropped.

The test flow is:
1. Run the SM on remote node,
2. Restart the driver.
4. Ping some destination,
3. Observe that first ICMP request will be dropped.

Fixes: fc791b633515 ("IB/ipoib: move back IB LL address into the hard
header")
Signed-off-by: Erez Shitrit <erezsh@mellanox.com>
Signed-off-by: Noa Osherovich <noaos@mellanox.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Tested-by: Yan Xu <jenny.x.xu@oracle.com>
Reviewed-by: Håkon Bugge <haakon.bugge@oracle.com>
Signed-off-by: Yuval Shaia <yuval.shaia@oracle.com>
8 years agomm: memcontrol: do not recurse in direct reclaim
Johannes Weiner [Fri, 28 Oct 2016 00:46:56 +0000 (17:46 -0700)]
mm: memcontrol: do not recurse in direct reclaim

On 4.0, we saw a stack corruption from a page fault entering direct
memory cgroup reclaim, calling into btrfs_releasepage(), which then
tried to allocate an extent and recursed back into a kmem charge ad
nauseam:

  [...]
  btrfs_releasepage+0x2c/0x30
  try_to_release_page+0x32/0x50
  shrink_page_list+0x6da/0x7a0
  shrink_inactive_list+0x1e5/0x510
  shrink_lruvec+0x605/0x7f0
  shrink_zone+0xee/0x320
  do_try_to_free_pages+0x174/0x440
  try_to_free_mem_cgroup_pages+0xa7/0x130
  try_charge+0x17b/0x830
  memcg_charge_kmem+0x40/0x80
  new_slab+0x2d9/0x5a0
  __slab_alloc+0x2fd/0x44f
  kmem_cache_alloc+0x193/0x1e0
  alloc_extent_state+0x21/0xc0
  __clear_extent_bit+0x2b5/0x400
  try_release_extent_mapping+0x1a3/0x220
  __btrfs_releasepage+0x31/0x70
  btrfs_releasepage+0x2c/0x30
  try_to_release_page+0x32/0x50
  shrink_page_list+0x6da/0x7a0
  shrink_inactive_list+0x1e5/0x510
  shrink_lruvec+0x605/0x7f0
  shrink_zone+0xee/0x320
  do_try_to_free_pages+0x174/0x440
  try_to_free_mem_cgroup_pages+0xa7/0x130
  try_charge+0x17b/0x830
  mem_cgroup_try_charge+0x65/0x1c0
  handle_mm_fault+0x117f/0x1510
  __do_page_fault+0x177/0x420
  do_page_fault+0xc/0x10
  page_fault+0x22/0x30

On later kernels, kmem charging is opt-in rather than opt-out, and that
particular kmem allocation in btrfs_releasepage() is no longer being
charged and won't recurse and overrun the stack anymore.

But it's not impossible for an accounted allocation to happen from the
memcg direct reclaim context, and we needed to reproduce this crash many
times before we even got a useful stack trace out of it.

Like other direct reclaimers, mark tasks in memcg reclaim PF_MEMALLOC to
avoid recursing into any other form of direct reclaim.  Then let
recursive charges from PF_MEMALLOC contexts bypass the cgroup limit.

Link: http://lkml.kernel.org/r/20161025141050.GA13019@cmpxchg.org
Orabug: 25430551

Signed-off-by: Johannes Weiner <hannes@cmpxchg.org>
Acked-by: Michal Hocko <mhocko@suse.com>
Cc: Vladimir Davydov <vdavydov.dev@gmail.com>
Cc: Tejun Heo <tj@kernel.org>
Cc: <stable@vger.kernel.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit 89a2848381b5fcd9c4d9c0cd97680e3b28730e31)
Signed-off-by: Ashish Samant <ashish.samant@oracle.com>
Signed-off-by: Nitin Gupta <nitin.m.gupta@oracle.com>
8 years agomemcg: ratify and consolidate over-charge handling
Tejun Heo [Fri, 6 Nov 2015 02:46:17 +0000 (18:46 -0800)]
memcg: ratify and consolidate over-charge handling

try_charge() is the main charging logic of memcg.  When it hits the limit
but either can't fail the allocation due to __GFP_NOFAIL or the task is
likely to free memory very soon, being OOM killed, has SIGKILL pending or
exiting, it "bypasses" the charge to the root memcg and returns -EINTR.
While this is one approach which can be taken for these situations, it has
several issues.

* It unnecessarily lies about the reality.  The number itself doesn't
  go over the limit but the actual usage does.  memcg is either forced
  to or actively chooses to go over the limit because that is the
  right behavior under the circumstances, which is completely fine,
  but, if at all avoidable, it shouldn't be misrepresenting what's
  happening by sneaking the charges into the root memcg.

* Despite trying, we already do over-charge.  kmemcg can't deal with
  switching over to the root memcg by the point try_charge() returns
  -EINTR, so it open-codes over-charing.

* It complicates the callers.  Each try_charge() user has to handle
  the weird -EINTR exception.  memcg_charge_kmem() does the manual
  over-charging.  mem_cgroup_do_precharge() performs unnecessary
  uncharging of root memcg, which BTW is inconsistent with what
  memcg_charge_kmem() does but not broken as [un]charging are noops on
  root memcg.  mem_cgroup_try_charge() needs to switch the returned
  cgroup to the root one.

The reality is that in memcg there are cases where we are forced and/or
willing to go over the limit.  Each such case needs to be scrutinized and
justified but there definitely are situations where that is the right
thing to do.  We alredy do this but with a superficial and inconsistent
disguise which leads to unnecessary complications.

This patch updates try_charge() so that it over-charges and returns 0 when
deemed necessary.  -EINTR return is removed along with all special case
handling in the callers.

While at it, remove the local variable @ret, which was initialized to zero
and never changed, along with done: label which just returned the always
zero @ret.

Orabug: 25430551

Signed-off-by: Tejun Heo <tj@kernel.org>
Reviewed-by: Vladimir Davydov <vdavydov@parallels.com>
Acked-by: Michal Hocko <mhocko@suse.com>
Acked-by: Johannes Weiner <hannes@cmpxchg.org>
Signed-off-by: Andrew Morton <akpm@linux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
(cherry picked from commit 10d53c748bc9531f47e13f98e32ef28be4399862)
Signed-off-by: Nitin Gupta <nitin.m.gupta@oracle.com@oracle.com>
Signed-off-by: Ashish Samant <ashish.samant@oracle.com>
8 years agonfs: Fix "Don't increment lock sequence ID after NFS4ERR_MOVED"
Chuck Lever [Thu, 26 Jan 2017 20:14:52 +0000 (15:14 -0500)]
nfs: Fix "Don't increment lock sequence ID after NFS4ERR_MOVED"

Lock sequence IDs are bumped in decode_lock by calling
nfs_increment_seqid(). nfs_increment_sequid() does not use the
seqid_mutating_err() function fixed in commit 059aa7348241 ("Don't
increment lock sequence ID after NFS4ERR_MOVED").

Fixes: 059aa7348241 ("Don't increment lock sequence ID after ...")
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Tested-by: Xuan Qi <xuan.qi@oracle.com>
Cc: stable@vger.kernel.org # v3.7+
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Orabug: 25416941
(cherry picked from commit 406dab8450ec76eca88a1af2fc15d18a2b36ca49)
Signed-off-by: Todd Vierling <todd.vierling@oracle.com>
Reviewed-by: Jack Vogel <jack.vogel@oracle.com>
8 years agonfs: Don't increment lock sequence ID after NFS4ERR_MOVED
Chuck Lever [Sun, 22 Jan 2017 19:04:29 +0000 (14:04 -0500)]
nfs: Don't increment lock sequence ID after NFS4ERR_MOVED

Xuan Qi reports that the Linux NFSv4 client failed to lock a file
that was migrated. The steps he observed on the wire:

1. The client sent a LOCK request to the source server
2. The source server replied NFS4ERR_MOVED
3. The client switched to the destination server
4. The client sent the same LOCK request to the destination
   server with a bumped lock sequence ID
5. The destination server rejected the LOCK request with
   NFS4ERR_BAD_SEQID

RFC 3530 section 8.1.5 provides a list of NFS errors which do not
bump a lock sequence ID.

However, RFC 3530 is now obsoleted by RFC 7530. In RFC 7530 section
9.1.7, this list has been updated by the addition of NFS4ERR_MOVED.

Reported-by: Xuan Qi <xuan.qi@oracle.com>
Signed-off-by: Chuck Lever <chuck.lever@oracle.com>
Cc: stable@vger.kernel.org # v3.7+
Signed-off-by: Trond Myklebust <trond.myklebust@primarydata.com>
Orabug: 25416941
(cherry picked from commit 059aa734824165507c65fd30a55ff000afd14983)
Signed-off-by: Todd Vierling <todd.vierling@oracle.com>
Reviewed-by: Jack Vogel <jack.vogel@oracle.com>
8 years agocrypto: mcryptd - Check mcryptd algorithm compatibility
tim [Mon, 5 Dec 2016 19:46:31 +0000 (11:46 -0800)]
crypto: mcryptd - Check mcryptd algorithm compatibility

Orabug: 25415629
CVE: CVE-2016-10147

Algorithms not compatible with mcryptd could be spawned by mcryptd
with a direct crypto_alloc_tfm invocation using a "mcryptd(alg)" name
construct.  This causes mcryptd to crash the kernel if an arbitrary
"alg" is incompatible and not intended to be used with mcryptd.  It is
an issue if AF_ALG tries to spawn mcryptd(alg) to expose it externally.
But such algorithms must be used internally and not be exposed.

We added a check to enforce that only internal algorithms are allowed
with mcryptd at the time mcryptd is spawning an algorithm.

Link: http://marc.info/?l=linux-crypto-vger&m=148063683310477&w=2
Cc: stable@vger.kernel.org
Reported-by: Mikulas Patocka <mpatocka@redhat.com>
Signed-off-by: Tim Chen <tim.c.chen@linux.intel.com>
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
(cherry picked from commit 48a992727d82cb7db076fa15d372178743b1f4cd)
Signed-off-by: Somasundaram Krishnasamy <somasundaram.krishnasamy@oracle.com>
Reviewed-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: bump version number
Jacob Keller [Thu, 3 Nov 2016 23:05:47 +0000 (16:05 -0700)]
fm10k: bump version number

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 6721f2dad51164fe9da7b3c2f8156529d716b888)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: do not clear global mailbox interrupt bits
Ngai-Mint Kwan [Wed, 2 Nov 2016 23:44:47 +0000 (16:44 -0700)]
fm10k: do not clear global mailbox interrupt bits

Partially revert commit 5e93cbadd3e9 ("fm10k: Reset mailbox global
interrupts", 2016-06-07)

The register bits related to this commit are now solely being handled by
the IES API. Recent changes in the IES API will allow an automatic
recovery from improper handling of these bits.

Signed-off-by: Ngai-Mint Kwan <ngai-mint.kwan@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit f0524955177c3e5cea89ceec56ad9d538530fe4f)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: request reset when mbx->state changes
Ngai-Mint Kwan [Wed, 2 Nov 2016 23:44:46 +0000 (16:44 -0700)]
fm10k: request reset when mbx->state changes

Multiple IES API resets can cause a race condition where the mailbox
interrupt request bits can be cleared before being handled. This can
leave certain mailbox messages from the PF to be untreated and the PF
will enter in some inactive state. If this situation occurs, the IES API
will initiate a mailbox version reset which, then, trigger a mailbox
state change. Once this mailbox transition occurs (from OPEN to CONNECT
state), a request for reset will be returned.

This ensures that PF will undergo a reset whenever IES API encounters an
unknown global mailbox interrupt event or whenever the IES API
terminates.

Signed-off-by: Ngai-Mint Kwan <ngai-mint.kwan@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 2f3fc1e6200309ccf87f61dea56e57e563c4f800)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: remove extraneous variable definition in fm10k_ethtool.c
Jacob Keller [Wed, 2 Nov 2016 23:44:45 +0000 (16:44 -0700)]
fm10k: remove extraneous variable definition in fm10k_ethtool.c

We don't need to typecast a u8 * into a char *, so just remove the extra
variable.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit aee243334461f83f3cb2e8c19e9a6cdedd35ee4b)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: wrap long line for alloc_workqueue
Jacob Keller [Thu, 25 Aug 2016 21:15:39 +0000 (14:15 -0700)]
fm10k: wrap long line for alloc_workqueue

Trivial change here to cleanup a checkpatch.pl warning that got
introduced when changing to alloc_workqueue.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
Note: removed unused get_ts_info routine (missed previous commit change)
(cherry picked from commit 5e3d033ec175b3d0a8da2b1fc9aee8bc8acf3cfc)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: use generic ethtool_op_get_ts_info callback
Jacob Keller [Thu, 25 Aug 2016 21:06:54 +0000 (14:06 -0700)]
fm10k: use generic ethtool_op_get_ts_info callback

This generic callback is for drivers which have software Tx timestamp
support enabled. Without this, PTP applications requesting software
timestamps may complain that the requested mode is not supported.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit bab02a69296682881cae280587618978c357e26e)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: don't re-map queues when a mailbox message suffices
Jacob Keller [Tue, 9 Aug 2016 00:08:19 +0000 (17:08 -0700)]
fm10k: don't re-map queues when a mailbox message suffices

When the PF assigns a new MAC address to a VF it uses the base address
registers to store the MAC address. This allows a VF which loads after
this setup the ability to get the initial address without having to wait
for a mailbox message. Unfortunately to do this, the PF must take queue
ownership away from the VF, which can cause fault errors when there is
already an active VF driver.

This queue ownership assignment causes race condition between the PF and
the VF such that potentially a VF can cause FUM fault errors due to
normal PF/VF driver behavior.

It is not safe to simply allow the PF to write the base address
registers without taking queue ownership back as the PF must also
disable the queues, and this would impact active VF use. The current
code is safe because the queue ownership will prevent the VF from
actually writing but does trigger the FUM fault.

We can do better by simply avoiding the register write process when
a mailbox message suffices. If the message can be sent over the mailbox,
then we will not perform the queue ownership assignment and we won't
update the base address to be the same as the MAC address.

We do still have to write the TXQCTL registers in order to update the
VID of the queue. This is necessary because the TXQCTL register is
read-only from the VF, and thus the VF cannot do this for itself. This
register does not need to wait for the Tx queue to be disabled and is
safe for the PF to write during normal VF operation, so we move this
write to the top of the function above the mailbox message. Without
this, the TXQCTL register would be misconfigured and cause the VF to Tx
hang.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 325782a173d1858bb67a827905e264cd128241d0)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: don't clear the RXQCTL register when enabling or disabling queues
Jacob Keller [Wed, 3 Aug 2016 22:05:27 +0000 (15:05 -0700)]
fm10k: don't clear the RXQCTL register when enabling or disabling queues

Ensure that other bits in the RXQCTL register do not get cleared. This
ensures that bits related to queue ownership are maintained.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit c689eff124cb231d91777a447fa05b30939da7b1)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: remove unnecessary extra parenthesis around ((~value))
Jacob Keller [Fri, 22 Jul 2016 23:00:37 +0000 (16:00 -0700)]
fm10k: remove unnecessary extra parenthesis around ((~value))

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 9717c7721302d217fb73f59eb3ab6314ffff54aa)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: don't try to stop queues if we've lost hw_addr
Jacob Keller [Thu, 23 Jun 2016 20:54:01 +0000 (13:54 -0700)]
fm10k: don't try to stop queues if we've lost hw_addr

In the event of a surprise remove, we expect the driver to go down,
which includes calling .stop_hw(). However, this function will return an
error because the queues won't appear to cleanly disable. Prevent this
and avoid the unnecessary checks by just returning when
FM10K_REMOVED(hw->hw_addr) is true.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 5f45c83024c49f71e50cf8156b165f11d2ab35a9)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: don't continue probe if PCI device not in normal IO state
Jacob Keller [Thu, 23 Jun 2016 20:31:01 +0000 (13:31 -0700)]
fm10k: don't continue probe if PCI device not in normal IO state

In the event of an uncorrectable AER error occurring when the driver has
not loaded, the recovery routines are not done. This is done because
future loads of the driver may not be aware of the IO state and may not
be able to recover at all. In this case, when we next load the driver it
fails due to what appears to be a surprise remove event. Instead, add
a check to ensure that the device is in the normal IO state before
continuing to probe. This allows us to give a more descriptive message
of what is wrong.

Without this change, the driver will attempt to probe up to our first
call of .reset_hw() which will be unable to read registers and act as if
a surprise remove event occurred.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 18095937cb1c66b8ff944de02cf04d1497008352)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: print error code when pci_enable_device_mem fails during probe
Jacob Keller [Thu, 23 Jun 2016 20:31:00 +0000 (13:31 -0700)]
fm10k: print error code when pci_enable_device_mem fails during probe

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 76ef0fc5a7519690a6a87a248f5709463f584292)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: NAPI polling routine must return actual work done
Jacob Keller [Mon, 20 Jun 2016 17:39:32 +0000 (10:39 -0700)]
fm10k: NAPI polling routine must return actual work done

When fm10k_poll fully cleans rings it returns 0. This is incorrect as it
messes up the budget accounting in the core NAPI code. Fix this by
returning actual work done, capped at budget - 1 since the core doesn't
expect a return of the full budget when the driver modifies the NAPI
status.

Cc: Paolo Abeni <pabeni@redhat.com>
Cc: Venkatesh Srinivas <venkateshs@google.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit e5fbfb78641ff0c5139ae665289ed9f91524265e)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: prefer READ_ONCE instead of ACCESS_ONCE
Jacob Keller [Fri, 17 Jun 2016 23:21:11 +0000 (16:21 -0700)]
fm10k: prefer READ_ONCE instead of ACCESS_ONCE

While technically not needed, as all our uses of ACCESS_ONCE are scalar
types, we already use READ_ONCE in a few places, and for code
readability we can swap all the uses of the older ACCESS_ONCE into
READ_ONCE.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit ce4dad2ce231aa5258ddfc98f8a80d958643c014)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: remove fm10k_get_reta_size from namespace
Jacob Keller [Fri, 17 Jun 2016 21:36:45 +0000 (14:36 -0700)]
fm10k: remove fm10k_get_reta_size from namespace

The function is only used in fm10k_ethtool.c, so make it static.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 88cdcfec9a46af47291207d86493e919093df3d1)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: use variadic form of alloc_workqueue
Jacob Keller [Thu, 9 Jun 2016 22:42:36 +0000 (15:42 -0700)]
fm10k: use variadic form of alloc_workqueue

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 4aa0bd54d4234d7bdc7cce331e7097c9393aca6e)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: use software values when checking for Tx hangs in hot path
Jacob Keller [Thu, 9 Jun 2016 21:56:05 +0000 (14:56 -0700)]
fm10k: use software values when checking for Tx hangs in hot path

A previous patch added support to check for hardware Tx pending in the
fm10k_down routine. This support was intended to ensure that we
accurately check what the hardware state is. However, checking for Tx
hangs in this manor during the hotpath results in a large performance
hit. Avoid this by making the hotpath check use the SW counters instead.

Fixes: a0f53cf49cb0 ("fm10k: use actual hardware registers when checking for pending Tx", 2016-06-08)
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 5b9e4432db038eefcafe2be468e3adaf2edbbe91)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: fix PCI device enable_cnt leak in .io_slot_reset
Jacob Keller [Thu, 9 Jun 2016 19:02:03 +0000 (12:02 -0700)]
fm10k: fix PCI device enable_cnt leak in .io_slot_reset

A previous patch removed the pci_disable_device() call in
.io_error_detected. This call corresponded to a pci_enable_device_mem()
call within .io_slot_reset handler. Change the call here to
a pci_reenable_device() so that it does not increment and leak the
enable_cnt reference count for the device. Without this change, VF
devices may fail during an unbind/bind, and we'll never zero the
reference counter for the pci_dev structure.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit e59a393d089d08a4622de07f941dd3629fcaec6a)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: bump version number
Jacob Keller [Tue, 7 Jun 2016 23:09:02 +0000 (16:09 -0700)]
fm10k: bump version number

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 5264cc63ba10ebfa0e54e3e641cce2656c7a60e8)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: return proper error code when pci_enable_msix_range fails
Jacob Keller [Tue, 7 Jun 2016 23:09:01 +0000 (16:09 -0700)]
fm10k: return proper error code when pci_enable_msix_range fails

The pci_enable_msix_range() function returns a positive value of the
number of allocated vectors if it succeeds. On failure it returns
a negative error code. Return this code properly so that the error
message printed by the driver will show the actual error code instead of
being masked by -ENOMEM.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 30e23b711c7b92dea664e61c3eb4b2cdc4b262e9)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: force link to remain down for at least a second on resume events
Jacob Keller [Tue, 7 Jun 2016 23:09:00 +0000 (16:09 -0700)]
fm10k: force link to remain down for at least a second on resume events

When we resume from an AER recovery with many active VFs, the PF sees
many spurious link up and link down events. Prevent this by delaying
link down for at least one second after the resume event.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 0356b23bcc347d4020f2883ad083ab54e573e214)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: implement request_lport_map pointer
Jacob Keller [Tue, 7 Jun 2016 23:08:59 +0000 (16:08 -0700)]
fm10k: implement request_lport_map pointer

If the fm10k interface is brought up, but the switch manager software is
not running, the driver will continuously request the lport map every
few seconds in the base driver watchdog routine. Eventually after
several minutes the switch mailbox Tx fifo will fill up and the mailbox
will timeout, resulting in a reset. This reset will appear as if for no
reason, and occurs regularly every few minutes until the switch manager
software is loaded.

Prevent this from happening by only requesting the lport map after we've
verified the switch mailbox is tx_ready. In order to simplify code logic
and reduce code duplication, implement this as a new function pointer
"mac.ops.request_lport_map" which the VF will not implement. Otherwise,
we have to duplicate the tx_ready check outside of
fm10k_get_host_state_generic, or re-implement most of
fm10k_get_host_state_generic in the pf version.

The resulting code is simpler and easier to understand, and prevents the
PF from continuously requesting lport map and filling the Tx fifo of
a switch mailbox that isn't ready.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 0afd20e5573c4aa976f1b935b6df73592b46ded5)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: check if PCIe link is restored
Jacob Keller [Tue, 7 Jun 2016 23:08:58 +0000 (16:08 -0700)]
fm10k: check if PCIe link is restored

Sometimes, a VF driver will lose PCIe address access, such as due to
a PF FLR event. In fm10k_detach_subtask, poll and check whether the
PCIe register space is active again and restore the device when it has.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 9d7dbf0604d37f8cae0c6d184a646037636a8b30)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: enable bus master after every reset
Jacob Keller [Tue, 7 Jun 2016 23:08:57 +0000 (16:08 -0700)]
fm10k: enable bus master after every reset

If an FLR occurs, VF devices will be knocked out of bus master mode, and
the driver will be unable to recover from the reset properly, resulting
in malicious driver events and an infinite reset loop. In the normal
case, the bus master mode will already be enabled and this call will
essentially be a no-op. Since we're doing this every reset, it is
possible we could remove the other calls to pci_set_master() but it
seems not harmful to just leave them in place.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 0d63a8f50e7a7810b0631cd8ea66d44dafb215e6)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: implement reset_notify handler for PCIe FLR events
Jacob Keller [Tue, 7 Jun 2016 23:08:55 +0000 (16:08 -0700)]
fm10k: implement reset_notify handler for PCIe FLR events

When a function level PCI reset is triggered using sysfs, it calls the
driver's .reset_notify error handler. Implement a handler based on the
now split fm10k_prepare_for_reset and fm10k_handle_reset functions, so
that we fully reset the driver when the PCI function level reset occurs.
This also ensures the reset is handled in a clean way by first disabling
all the driver bits first and then restoring them after the function
reset. Previously the stack simply performed a blind function reset and
our driver didn't take any part in the process.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 0593186a17d8427804be162a724a1576367e4d4c)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: implement prepare_suspend and handle_resume
Jacob Keller [Tue, 7 Jun 2016 23:08:53 +0000 (16:08 -0700)]
fm10k: implement prepare_suspend and handle_resume

Implement fm10k_prepare_suspend and fm10k_handle_resume functions which
abstract around the now existing fm10k_prepare_for_reset and
fm10k_handle_reset. The new functions also handle stopping the service
task, which is something that the original re-init flow does not need.

Every other location that does a suspend/resume type flow is expected to
use these functions, because otherwise they may have conflicts with the
running watchdog routines. This also has the effect of preventing
possible surprise remove events during handling of FLR events and PCIe
errors.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit dc4b76c0fe1753940b15413ae5fb9829552824ec)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: split fm10k_reinit into two functions
Jacob Keller [Tue, 7 Jun 2016 23:08:52 +0000 (16:08 -0700)]
fm10k: split fm10k_reinit into two functions

There are several flows in the driver which perform the similar function
of tearing down software and restoring software to recover from certain
errors or PCIe events, including:

  * fm10k_reinit
  * fm10k_suspend/resume
  * fm10k_io_error_detected/fm10k_io_resume

In addition, we want to implement a .reset_notify() handler as well
which will also perform similar function.

Rework how the driver codes reset and resume flows by separating out the
reinit logic into two functions "fm10k_prepare_for_reset" and
"fm10k_handle_reset". This first step will allow us to re-use this
functionality in the similar blocks of code instead of re-coding the
same sequence of events slightly different.

The end result should be more maintainable and correct, fixing several
inconsistencies with the work flow.

The new functions expect to take the rtnl_lock() themselves, and it does
have the unfortunate side effect of having the reinit flow take then
release then take the rtnl_lock. However, this minor downside is
out weighted by the benefits of code reduction and reducing needless
difference between these flows.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 40de1fad417cb7c6958c214ec1f4c8c29101cb7f)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: wait for queues to drain if stop_hw() fails once
Jacob Keller [Tue, 7 Jun 2016 23:08:51 +0000 (16:08 -0700)]
fm10k: wait for queues to drain if stop_hw() fails once

It turns out that sometimes during a reset the Tx queues will be
temporarily stuck longer than .stop_hw() expects. Work around this issue
by attempting to .stop_hw() first. If it tails, wait a number of
attempts until the Tx queues appear to be drained. After this, attempt
stop_hw() again. This ensures that we avoid waiting if we don't need to,
such as during the first initialization of a VF, and give the proper
amount of time necessary to recover from most situations. It is possible
that the hardware is actually stuck. For PFs, this is usually fixed by
a datapath reset. Unfortunately the VF cannot request a similar reset
for itself.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 94877768cfaa99f7b3757f29632faa5945f18872)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: only warn when stop_hw fails with FM10K_ERR_REQUESTS_PENDING
Jacob Keller [Tue, 7 Jun 2016 23:08:50 +0000 (16:08 -0700)]
fm10k: only warn when stop_hw fails with FM10K_ERR_REQUESTS_PENDING

When stop_hw() routine fails with FM10K_ERR_REQUESTS_PENDING, this
indicates that the Tx or Rx queues did not shutdown within the time
limit. Print a more suitable message at the dev_info level instead of
dev_err.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 106ca42356b49a1ae6199e6630ec40df82ff7421)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: use actual hardware registers when checking for pending Tx
Jacob Keller [Tue, 7 Jun 2016 23:08:49 +0000 (16:08 -0700)]
fm10k: use actual hardware registers when checking for pending Tx

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 34bad71c7c25840cd78dff12d75497d3aab5544d)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: perform data path reset even when switch is not ready
Jacob Keller [Tue, 7 Jun 2016 23:08:48 +0000 (16:08 -0700)]
fm10k: perform data path reset even when switch is not ready

A while ago, an additional check for the switch being ready was added to
reset_hw. A recent refactor accidentally made this check return an error
code on failure which caused fm10k_probe to fail when the switch wasn't
brought up first. The original reasoning for the check was to prevent
additional data path reset when the fabric wasn't ready yet. However,
there isn't a compelling reason to keep the check, as the data path
reset will restore hardware to a known good state. Remove the check and
perform the data path reset regardless of the switch manager state.

An alternative fix is to return FM10K_SUCCESS instead, and bypass the
actual data path reset. This should be fine as we will perform
a reset_hw once the switch is active. However, since data path reset
will reset many parts of the hardware it seems better to just perform
the reset regardless of switch state.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 892c9e0872bea23ac2dc774ea950e6526f157bd5)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: don't stop reset due to FM10K_ERR_REQUESTS_PENDING
Jacob Keller [Tue, 7 Jun 2016 23:08:47 +0000 (16:08 -0700)]
fm10k: don't stop reset due to FM10K_ERR_REQUESTS_PENDING

Don't report FM10K_ERR_REQUESTS_PENDING when we fail to disable queues
within the timeout. This can occur due to a hardware Tx hang, or when
the switch ethernet fabric is resetting while we are transmitting
traffic. It can sometimes take up to 500ms before the Tx DMA engine
gives up. Instead, just skip the DMA engine check and perform
a data-path reset anyways. Add a statistic counter to keep track of the
number of resets occurring while we have pending DMA on the rings.

In order to prevent having to re-assign err to 0, re-order the
last few items of the reset_hw_pf function so that we don't perform
"return err" at the end.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit ce33624f37f43267e5fa0810a058807d268142fb)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: Reset mailbox global interrupts
Ngai-Mint Kwan [Tue, 7 Jun 2016 23:08:46 +0000 (16:08 -0700)]
fm10k: Reset mailbox global interrupts

When a data path reset is initiated, write control to the PCIE_GMBX is
yanked from the switch manager. The switch manager writes to this
register to clear mailbox global interrupt bits as part of its mailbox
interrupt handling routine. When the device recovers from the data path
reset and these bits are not cleared, it will prevent future mailbox
global interrupts from being triggered. Upon confirming that the device
has exited from a data path reset, clear these bits to ensure the proper
functioning of the mailbox global interrupt.

Signed-off-by: Ngai-Mint Kwan <ngai-mint.kwan@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 5e93cbadd3e9db342e48582178b7d241097f51c4)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: prevent multiple threads updating statistics
Jacob Keller [Tue, 7 Jun 2016 23:08:45 +0000 (16:08 -0700)]
fm10k: prevent multiple threads updating statistics

Also prevent updating stats while the interface is down. If we're
already updating stats, just return doing nothing. When we take the
device down, block stat updates until we come back up. This ensures that
we avoid tearing down rings when we're updating statistics, and prevents
updating statistics until we're up.

We can't re-use the __FM10K_DOWN for this because it wouldn't prevent
multiple threads from accessing statistics. Neither does it prevent the
case where we start updating stats and then start going down in another
thread.

The fm10k_get_stats64 is except from this, because it has a completely
different flow which does not suffer from the same issues as
fm10k_update_stats might.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 9d73edee5900baff47887b36e9b97f45a7d13b6d)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: avoid possible null pointer dereference in fm10k_update_stats
Jacob Keller [Fri, 3 Jun 2016 22:42:12 +0000 (15:42 -0700)]
fm10k: avoid possible null pointer dereference in fm10k_update_stats

It's currently possible for fm10k_update_stats to be called during the
window when we go down and the rings are removed. This can result in
a null pointer dereference. In fm10k_get_stats64 we work around this by
using ACCESS_ONCE and a null pointer check inside the loop. Use this
same flow in the fm10k_update_stats to avoid the potential null pointer.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit b624714bc90064eeefd9ba7564e90865eef00421)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: no need to continue in fm10k_down if __FM10K_DOWN already set
Jacob Keller [Fri, 3 Jun 2016 22:42:11 +0000 (15:42 -0700)]
fm10k: no need to continue in fm10k_down if __FM10K_DOWN already set

Return early from fm10k_down() when we are already down, since that
means another thread is either already finished or has started going
down, so shouldn't conflict with them.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 1b00c6c064302354fce71d7c363945fe9e967f7c)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: Remove create_workqueue
Bhaktipriya Shridhar [Wed, 1 Jun 2016 15:40:09 +0000 (21:10 +0530)]
fm10k: Remove create_workqueue

alloc_workqueue replaces deprecated create_workqueue().

A dedicated workqueue has been used since the workitem (viz
fm10k_service_task, which manages and runs other subtasks) is involved in
normal device operation and requires forward progress under memory
pressure.

create_workqueue has been replaced with alloc_workqueue with max_active
as 0 since there is no need for throttling the number of active work
items.

Since network devices may be used in memory reclaim path,
WQ_MEM_RECLAIM has been set to guarantee forward progress.

flush_workqueue is unnecessary since destroy_workqueue() itself calls
drain_workqueue() which flushes repeatedly till the workqueue
becomes empty. Hence the call to flush_workqueue() has been dropped.

Signed-off-by: Bhaktipriya Shridhar <bhaktipriya96@gmail.com>
Acked-by: Tejun Heo <tj@kernel.org>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 0a38c17a21a0965b4853211afa1d3e85428e6170)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: fix incorrect index calculation in fm10k_write_reta
Jacob Keller [Mon, 18 Apr 2016 22:45:00 +0000 (15:45 -0700)]
fm10k: fix incorrect index calculation in fm10k_write_reta

The index calculated when looping through the indir array passed to
fm10k_write_reta was incorrectly calculated as the first part i needs to
be multiplied by 4.

Fixes: 0cfea7a65738 ("fm10k: fix possible null pointer deref after kcalloc", 2016-04-13)
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 34875887f360d7bd0b7f0a89f7c6d65eca616ee3)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: Align Rx buffers to 512B blocks
Alexander Duyck [Fri, 15 Apr 2016 17:00:46 +0000 (13:00 -0400)]
fm10k: Align Rx buffers to 512B blocks

While reviewing the i40e driver changes to support page based receive I
realized that I had overlooked the fact that the fm10k hardware required a
512 byte alignment for Rx buffers.  This patch is meant to address that by
changing the alignment for Rx buffers to 512 bytes instead of allowing it
to be L1 cache aligned.

Signed-off-by: Alexander Duyck <aduyck@mirantis.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit fb5677aa26a2ae2c94ffb889ab6802a42bf54b0b)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: don't use BIT() macro where the value isn't a bitmask
Jacob Keller [Thu, 14 Apr 2016 20:17:27 +0000 (13:17 -0700)]
fm10k: don't use BIT() macro where the value isn't a bitmask

The FM10K_MAX_DATA_PER_TXD is really just using a bitshift as a power of
2 operation in an efficient manner. We shouldn't represent this as a BIT()
because that obscures the intention of the operation.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 124579de462e6c232c4c321c3ec1ed81964cf78b)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: fix incorrect IPv6 extended header checksum
Jacob Keller [Thu, 7 Apr 2016 15:52:53 +0000 (08:52 -0700)]
fm10k: fix incorrect IPv6 extended header checksum

Check for and handle IPv6 extended headers so that Tx checksum offload
can be done. Also use skb_checksum_help for unexpected cases. This was
originally discovered in ixgbe.

Reported-by: Mark Rustad <mark.d.rustad@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit dc1b4c2b88b976a7882922e55666b20e28477c57)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: consistently use Intel(R) for driver names
Jacob Keller [Thu, 7 Apr 2016 15:21:21 +0000 (08:21 -0700)]
fm10k: consistently use Intel(R) for driver names

Update every header file and other locations to consistently use
Intel(R) instead of just Intel. Also update copyright year of files
which we modified.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 86641094678a90af278d1f44c0e47f817c9ba46e)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: fix possible null pointer deref after kcalloc
Jacob Keller [Thu, 7 Apr 2016 15:21:20 +0000 (08:21 -0700)]
fm10k: fix possible null pointer deref after kcalloc

When writing a new default redirection table, we needed to populate
a new RSS table using ethtool_rxfh_indir_default. We populated this
table into a region of memory allocated using kcalloc, but never checked
this for NULL. Fix this by moving the default table generation into
fm10k_write_reta. If this function is passed a table, use it. Otherwise,
generate the default table using ethtool_rxfh_indir_default, 4 at at
time.

Fixes: 0ea7fae44094 ("fm10k: use ethtool_rxfh_indir_default for default redirection table")
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 540a5d859010a239a99aba02a9fed7b255c0033e)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: Reset multicast mode when deleting lport
Ngai-Mint Kwan [Fri, 1 Apr 2016 23:17:39 +0000 (16:17 -0700)]
fm10k: Reset multicast mode when deleting lport

Deleting lport when multicast mode is configured to
FM10K_XCAST_MODE_ALLMULTI or FM10K_XCAST_MODE_PROMISC will result in
generating orphaned multicast-group entries in the switch manager.
Before deleting the lport, reset multicast mode to FM10K_XCAST_MODE_NONE
to flush out these multicast-group entries.

Signed-off-by: Ngai-Mint Kwan <ngai-mint.kwan@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 11ec36a974f59c99e8a4ff7040026569a43ab567)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: update comment regarding reserved bits check
Jacob Keller [Fri, 1 Apr 2016 23:17:38 +0000 (16:17 -0700)]
fm10k: update comment regarding reserved bits check

The original comment may be read incorrectly as referring to checking
the *entire* length is zero. However, it merely checks only the reserved
bits of both length and reserved in a small amount of code. Update the
comment to indicate this is a clever trick and clearly spell out that it
only checks the reserve bits.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit fb6515c8f03bbfdc99cff156becd5e14df1dd601)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: use different name than FM10K_VLAN_CLEAR for override bit
Jacob Keller [Fri, 1 Apr 2016 23:17:37 +0000 (16:17 -0700)]
fm10k: use different name than FM10K_VLAN_CLEAR for override bit

Use a new #define FM10K_VLAN_OVERRIDE even though we're using the exact
same bit. The reason for this is clarity in the code, otherwise you can
read FM10K_VLAN_CLEAR and think it should be removed. Also add a comment
explaining why the FM10K_VLAN_OVERRIDE bit is set.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 5c69df8a33408c82ac633c521be0acf71a690d43)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: use 8bit notation instead of 10bit notation for diagram
Jacob Keller [Fri, 1 Apr 2016 23:17:36 +0000 (16:17 -0700)]
fm10k: use 8bit notation instead of 10bit notation for diagram

The diagram represents bit layout of the multi-bit VLAN update message
format. Typically these diagrams are drawn using some power of 2 as the
base, to more easily grasp where fields split. Although the numbers
above can make it somewhat easy to understand which bit you're looking
at, it makes the break points not line up. Re-draw the numbers using
base 8, and mark the bit values every 8 bits at the top. This should
make it more easy to grasp the table quickly.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit d057d9a9446636293b4884d1a0da6ad5a7ef4e13)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: fix documentation of fm10k_tlv_parse_attr
Jacob Keller [Fri, 1 Apr 2016 23:17:35 +0000 (16:17 -0700)]
fm10k: fix documentation of fm10k_tlv_parse_attr

fm10k_tlv_parse_attr is supposed to return FM10K_NOT_IMPLEMENTED for any
TLV who's attribute id lies outside the range of results. It does not do
this today. In addition, the documentation does not indicate that other
attributes which are not implemented for a given TLV will be silently
ignored. Fix this. Clean up the logic so that we don't rely on the fact
that FM10K_NOT_IMPLEMENTED is greater than zero, as this can easily
cause confusion.

A future extension could look into some way of reporting unknown TLVs
in order to make issues more easily discoverable. We can't just return
FM10K_NOT_IMPLEMENTED here because we don't want to drop the entire
message if it has an unknown TLV.

While here, update the copyright year.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 4e160f2a59cec8f705583edfaa11ce5f3b3ef4a6)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: do not disable PCI device in fm10k_io_error_detected
Jacob Keller [Fri, 1 Apr 2016 23:17:34 +0000 (16:17 -0700)]
fm10k: do not disable PCI device in fm10k_io_error_detected

fm10k_io_error_detected() does not need to call pci_disable_device(). In
the cases where the reset needs to occur, the stack flow will result in
calling fm10k_remove() which already disables the PCI device. If we
leave the pci_disable_device(), we result in a warning about disabling
an already disabled device.

Many PCI drivers do call pci_disable_device() in their .error_detected()
routines, but it does not appear to be required. In addition, these
drivers have a check "is_pci_enabled()" call in their remove routines,
which is how they chose to handle the duplicate device disable.

This seems incorrect, since the PCI device structure is reference
counted. It is very possible that the reference count for the PCI device
could be greater than 1. In this case, you would remove the PCI device
within the error_detected routine, reducing count to 1, then remove it
again in the remove function, reducing it to zero. This would result in
yet another disable somewhere else failing. Thus, we shouldn't be using
is_pci_enabled() to check for this issue. Instead, just remove the
extraneous pci_device_disable() found within the error_detected routine.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 3417415c3a86d6bae8bfee495ce634f4d24e16b8)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: correctly handle LPORT_MAP error
Jacob Keller [Fri, 1 Apr 2016 23:17:33 +0000 (16:17 -0700)]
fm10k: correctly handle LPORT_MAP error

Currently, any error responses from the switch manager after an
LPORT_MAP request are silently ignored. At most the mailbox message will
be reported as an error. This can result in unexpected behavior when the
switch manager has configured a port with zero bandwidth. Add support
for reading the fm10k_swapi_error structure from LPORT_MAP responses.

If the message contains the necessary TLV and has a non-zero error code,
report link down, clear the dglort_map, and delay the next
get_host_state call by a reasonable delay. Also log an error message
indicating that the LPORT_MAP request failed.

The delay ensures preventing an interrupt storm on the switch manager,
and reduces the number of mailbox messages we send in this scenario
drastically.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit a7a7783adabc3cc7599f7dbf97fcc3b0d44087b7)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: Fix multicast mode sync issues
Ngai-Mint Kwan [Fri, 1 Apr 2016 23:17:32 +0000 (16:17 -0700)]
fm10k: Fix multicast mode sync issues

Multicast mode checking is no longer a requirement to perform unicast
and multicast address syncs. Specifically, a device operating in
promiscuous and/or all multicast mode is not excluded. The issue occurs
when the netdev is pre-configured to either multicast mode and is
enabled for the first time. The multicast-group table in the Switch
Manager will be missing obvious multicast entries associated to this
netdev.

Changes were also made to disallow unicast and multicast syncs with
VLAN 0. The Switch Manager considers VLAN 0 to be an invalid entry.
Requests with VLAN 0 by the netdev are only generated when the driver is
freshly installed and the default VID is not set.

Signed-off-by: Ngai-Mint Kwan <ngai-mint.kwan@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 8998763a7b57583ef2e07f68ea6a7d05bcfc1cfa)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: prevent RCU issues during AER events
Jacob Keller [Fri, 11 Mar 2016 17:52:32 +0000 (09:52 -0800)]
fm10k: prevent RCU issues during AER events

During an AER action response, we were calling fm10k_close without
holding the rtnl_lock() which could lead to possible RCU warnings being
produced due to 64bit stat updates among other causes. Similarly, we
need rtnl_lock() around fm10k_open during fm10k_io_resume. Follow the
same pattern elsewhere in the driver and protect the entire open/close
sequence.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 1e4c32f3ede19bdb738aec2cc3cf69439d7b9310)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: use DRV_SUMMARY to reduce code duplication
Jacob Keller [Thu, 10 Mar 2016 00:36:08 +0000 (16:36 -0800)]
fm10k: use DRV_SUMMARY to reduce code duplication

Use DRV_SUMMARY, similar to DRV_VERSION so that we don't have to
duplicate the driver summary in multiple places.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 2d0f76bedbddaacc465c7a3ebdc9f8c13f68d931)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: remove debug-statistics support
Jacob Keller [Fri, 4 Mar 2016 23:37:48 +0000 (15:37 -0800)]
fm10k: remove debug-statistics support

This change fixes an (ab)use of the ethtool stats API, which could
result in corrupt memory or misleading stat output. The ethtool stats
API is not robust enough to handle varying number of statistics due to
how it requests the size and allocates memory. Remove the poorly conceived
support originally added for extra debug statistics. In the future,
a new stats API may open up the ability to display these statistics.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 3ef2f563267892230681b1b8890d8f759d39e64d)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: add helper functions to set strings and data for ethtool stats
Jacob Keller [Fri, 1 Apr 2016 18:15:09 +0000 (11:15 -0700)]
fm10k: add helper functions to set strings and data for ethtool stats

Reduce duplicate code and the amount of indentation by adding
fm10k_add_stat_strings and fm10k_add_ethtool_stats functions which help
add fm10k_stat structures to the ethtool stats callbacks. This helps
increase ease of use for future stat additions, and increases code
readability.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 09401ae25191039f4aa45c13718595f550745c68)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: fix multi-bit VLAN update requests from VF
Jacob Keller [Thu, 31 Mar 2016 16:52:30 +0000 (09:52 -0700)]
fm10k: fix multi-bit VLAN update requests from VF

The VF uses a multi-bit update request to clear unused VLANs whenever it
resets. However, an accident in a previous refector broke multi-bit
updates for VFs, due to misreading a comment in fm10k_vf.c and
attempting to reduce code duplication. The problem occurs because
a multi-bit request has a non-zero length, and the PF would simply drop
any request with the upper 16 bits set.

We can't simply remove the check of the upper 16 bits and the call to
fm10k_iov_select vid, because this would remove the checks for default
VID and for ensuring no other VLANs can be enabled except pf_vid when it
has been set. To resolve that issue, this revision uses the
iov_select_vid when we have a single-bit update, and denies any
multi-bit update when the VLAN was administratively set by the PF. This
should be ok since the PF properly updates VLAN_TABLE when it assigns
the PF vid. This ensures that requests to add or remove the PF vid work
as expected, but a rogue VF could not use the multi-bit update as
a loophole to attempt receiving traffic on other VLANs.

Reported-by: Ngai-Mint Kwan <ngai-mint.kwan@intel.com>
Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit f808c5dbcdc393be8e9f676c61baac6a3db382c1)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: use ethtool_rxfh_indir_default for default redirection table
Jacob Keller [Wed, 17 Feb 2016 00:19:24 +0000 (16:19 -0800)]
fm10k: use ethtool_rxfh_indir_default for default redirection table

The fm10k driver used its own code for generating a default indirection
table on device load, which was not the same as the default generated by
ethtool when indir_size of 0 is passed to SRXFH. Take advantage of
ethtool_rxfh_indir_default() and simplify code to write the redirection
table to reduce some code duplication.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 0ea7fae44094b4ca06ea68105457a7dc64041bd3)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: fix a minor typo in some comments
Jacob Keller [Wed, 10 Feb 2016 22:45:51 +0000 (14:45 -0800)]
fm10k: fix a minor typo in some comments

s/funciton/function to resolve a typo, and cleanup grammar on a few
comments regarding processing the VF mailboxes.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit d8ec92f2cdcc7f2d06dd0a40b600b6da7d9d1070)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: correctly clean up when init_queueing_scheme fails
Jacob Keller [Wed, 10 Feb 2016 22:45:50 +0000 (14:45 -0800)]
fm10k: correctly clean up when init_queueing_scheme fails

Fix a kernel panic that occurs during surprise removal. Clear the
interface queue counts upon fm10k_init_msix_capability failure. This
prevents further code (fm10k_update_stats etc.) from attempting to
access unallocated queue vector or ring memory.

[  628.692648] BUG: unable to handle kernel NULL pointer dereference at 0000000000000068
[  628.692805] IP: [<ffffffffa0475caf>] fm10k_update_stats+0x7f/0x2c0 [fm10k]
[  628.693173] PGD 0
[  628.693759] Oops: 0000 [#1] SMP
[  628.699321] CPU: 10 PID: 8164 Comm: kworker/10:0 Tainted: G           OE  ------------   3.10.0-327.el7.x86_64 #1
[  628.700096] Hardware name: Supermicro X9DAi/X9DAi, BIOS 3.2 05/09/2015
[  628.700894] Workqueue: pciehp-1 pciehp_power_thread
[  628.701686] task: ffff88086559c500 ti: ffff8808593c0000 task.ti: ffff8808593c0000
[  628.702493] RIP: 0010:[<ffffffffa0475caf>]  [<ffffffffa0475caf>] fm10k_update_stats+0x7f/0x2c0 [fm10k]
[  628.703310] RSP: 0018:ffff8808593c3b00  EFLAGS: 00010282
[  628.704132] RAX: 0000000000000000 RBX: ffff880860760000 RCX: 0000000000000000
[  628.704963] RDX: ffff880860760b08 RSI: 0000000000000000 RDI: 0000000000000000
[  628.705794] RBP: ffff8808593c3b40 R08: 0000000000000000 R09: 0000000000000000
[  628.706604] R10: 0000000000000000 R11: ffff880860760c40 R12: 0000000000000080
[  628.707420] R13: ffff8808607608c0 R14: ffff880860779ec0 R15: ffff880860779f40
[  628.708238] FS:  0000000000000000(0000) GS:ffff88086f000000(0000) knlGS:0000000000000000
[  628.709071] CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
[  628.709923] CR2: 0000000000000068 CR3: 000000000194a000 CR4: 00000000001407e0
[  628.710752] DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
[  628.711596] DR3: 0000000000000000 DR6: 00000000ffff0ff0 DR7: 0000000000000400
[  628.712438] Stack:
[  628.713255]  ffff880860764458 ffff8808607608c0 ffff880860760000 ffff880860760000
[  628.714088]  0000000000000080 ffff8808607608c0 ffff880860779ec0 ffff880860779f40
[  628.714925]  ffff8808593c3b88 ffffffffa04780c5 ffff880860764458 0000000a8163cb5b
[  628.715752] Call Trace:
[  628.716560]  [<ffffffffa04780c5>] fm10k_down+0x155/0x1f0 [fm10k]
[  628.717367]  [<ffffffffa0479958>] fm10k_close+0x28/0xd0 [fm10k]
[  628.718184]  [<ffffffff81526365>] __dev_close_many+0x85/0xd0
[  628.718986]  [<ffffffff815264d8>] dev_close_many+0x98/0x120
[  628.719764]  [<ffffffff81527ab8>] rollback_registered_many+0xa8/0x230
[  628.720527]  [<ffffffff81527c80>] rollback_registered+0x40/0x70
[  628.721294]  [<ffffffff81529198>] unregister_netdevice_queue+0x48/0x80
[  628.722052]  [<ffffffff815291ec>] unregister_netdev+0x1c/0x30
[  628.722816]  [<ffffffffa04762b8>] fm10k_remove+0xd8/0xe0 [fm10k]
[  628.723581]  [<ffffffff81328c7b>] pci_device_remove+0x3b/0xb0
[  628.724340]  [<ffffffff813f5fbf>] __device_release_driver+0x7f/0xf0
[  628.725088]  [<ffffffff813f6053>] device_release_driver+0x23/0x30
[  628.725814]  [<ffffffff81321fe4>] pci_stop_bus_device+0x94/0xa0
[  628.726535]  [<ffffffff813220d2>] pci_stop_and_remove_bus_device+0x12/0x20
[  628.727249]  [<ffffffff8133de40>] pciehp_unconfigure_device+0xb0/0x1b0
[  628.727964]  [<ffffffff8133d822>] pciehp_disable_slot+0x52/0xd0
[  628.728664]  [<ffffffff8133d98a>] pciehp_power_thread+0xea/0x150
[  628.729358]  [<ffffffff8109d5fb>] process_one_work+0x17b/0x470
[  628.730036]  [<ffffffff8109e3cb>] worker_thread+0x11b/0x400
[  628.730730]  [<ffffffff8109e2b0>] ? rescuer_thread+0x400/0x400
[  628.731385]  [<ffffffff810a5aef>] kthread+0xcf/0xe0
[  628.732036]  [<ffffffff810a5a20>] ? kthread_create_on_node+0x140/0x140
[  628.732674]  [<ffffffff81645858>] ret_from_fork+0x58/0x90
[  628.733289]  [<ffffffff810a5a20>] ? kthread_create_on_node+0x140/0x140
[  628.733883] Code: 83 e8 01 48 8d 97 40 02 00 00 45 31 c0 4c 8d 9c c7 48 02 0
[  628.735202] RIP  [<ffffffffa0475caf>] fm10k_update_stats+0x7f/0x2c0 [fm10k]
[  628.735732]  RSP <ffff8808593c3b00>
[  628.736285] CR2: 0000000000000068
[  628.736846] ---[ end trace 9156088b311aff42 ]---

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 4be37c42a40cec94b0381b42e5796d3316f96c32)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: prevent possibly uninitialized variable
Bruce Allan [Wed, 10 Feb 2016 22:45:47 +0000 (14:45 -0800)]
fm10k: prevent possibly uninitialized variable

If 'attr_flag < (1 << (2 * FM10K_TEST_MSG_NESTED))' is ever false, err
will be used uninitialized.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit c4114e3db6429c665adc3db871685c474a467efe)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: add helper functions to set strings and data for ethtool stats
Jacob Keller [Fri, 5 Feb 2016 18:43:08 +0000 (10:43 -0800)]
fm10k: add helper functions to set strings and data for ethtool stats

Reduce duplicate code and the amount of indentation by adding
fm10k_add_stat_strings and fm10k_add_ethtool_stats functions which help
add fm10k_stat structures to the ethtool stats callbacks. This helps
increase ease of use for future stat additions, and increases code
readability. Skip handling of the per-queue stats as these will be
reworked in a following patch.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit d2e0721b18f320232dc36a0e4cc7beb620e8c9bd)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: free MBX IRQ before clearing interrupt scheme
Jacob Keller [Thu, 4 Feb 2016 18:47:58 +0000 (10:47 -0800)]
fm10k: free MBX IRQ before clearing interrupt scheme

During fm10k_io_error_detected we were clearing the interrupt scheme
before we freed the MBX IRQ. This causes a kernel panic because the MBX
IRQ are assigned after MSI-X initialization. Clearing the interrupt
scheme results in removing the MSI-X entry table. Fix this by freeing
the MBX IRQ before we clear the interrupt scheme, as we do elsewhere in
the driver.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit c8ed563bebeabbf0b1085b52916dd2fb6e219276)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: print error message when stop_hw fails
Jacob Keller [Thu, 4 Feb 2016 18:47:57 +0000 (10:47 -0800)]
fm10k: print error message when stop_hw fails

fm10k_stop_hw_generic calls fm10k_disable_queues_generic, which may
return an error code indicating that the queues were not stopped within
the time limit. Notify the user by displaying a message in the kernel
message ring, in a similar way to how we notify the user when reset_hw
fails. There isn't much we can do to recover from this error, so
currently nothing else is done.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 61e0217e83353cf895f8b2d0a187804171d119ca)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: base queue scheme covered by RSS
Jacob Keller [Thu, 4 Feb 2016 18:47:56 +0000 (10:47 -0800)]
fm10k: base queue scheme covered by RSS

In fm10k_set_num_queues, we previously assigned the base template. This
would always be overwritten by either fm10k_set_qos_queues or
fm10k_set_rss_queues. In either case, we don't need the base values, so
we can just remove them.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit b3525696adba1ecddff3d667680461cc533e63a4)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: don't initialize service task until later in probe
Jacob Keller [Thu, 4 Feb 2016 18:47:55 +0000 (10:47 -0800)]
fm10k: don't initialize service task until later in probe

Delay initialization of the service timer and service task until late
probe. If we don't wait, failures in probe do not properly cleanup the
service timer or service task items, which results in the kernel panic
below, potentially freezing the whole system. In addition, ensure that
the SERVICE_DISABLE bit is set before we request the MBX IRQ since the
MBX interrupt attempts to schedule the service task otherwise. This
prevents a similar trace from occurring after this change.

We didn't notice this issue before because probe almost always completes
successfully. I discovered it due to a mis-ordered mailbox handler
array, which resulted in the following failure when requesting mailbox
interrupt.

[  555.325619] ------------[ cut here ]------------
[  555.325628] WARNING: CPU: 0 PID: 4941 at lib/list_debug.c:33 __list_add+0xa0/0xd0()
[  555.325631] list_add corruption. prev->next should be next (ffffffff81f46648), but was           (null). (prev=ffff8807fad5d0e8).
<snip>
[  555.325722] CPU: 0 PID: 4941 Comm: insmod Tainted: G           OE   4.0.4-303.fc22.x86_64 #1
[  555.325725] Hardware name: Intel Corporation S2600CO/S2600CO, BIOS SE5C600.86B.02.03.8x23.060520140825 06/05/2014
[  555.325727]  0000000000000000 00000000b4f161b3 ffff88081a21f8e8 ffffffff81783124
[  555.325734]  0000000000000000 ffff88081a21f940 ffff88081a21f928 ffffffff8109c66a
[  555.325740]  0000000064000000 ffff8807fad5d0e8 ffff8807fad5d0e8 ffffffff81f46648
[  555.325746] Call Trace:
[  555.325752]  [<ffffffff81783124>] dump_stack+0x45/0x57
[  555.325757]  [<ffffffff8109c66a>] warn_slowpath_common+0x8a/0xc0
[  555.325759]  [<ffffffff8109c6f5>] warn_slowpath_fmt+0x55/0x70
[  555.325763]  [<ffffffff813ba270>] __list_add+0xa0/0xd0
[  555.325768]  [<ffffffff81102d1d>] __internal_add_timer+0x9d/0x110
[  555.325771]  [<ffffffff81102dbf>] internal_add_timer+0x2f/0xc0
[  555.325774]  [<ffffffff81104e5a>] mod_timer+0x12a/0x230
[  555.325782]  [<ffffffffa03d54ca>] fm10k_probe+0x69a/0xc80 [fm10k]
[  555.325787]  [<ffffffff813e8355>] local_pci_probe+0x45/0xa0
[  555.325791]  [<ffffffff8129cf42>] ? sysfs_do_create_link_sd.isra.2+0x72/0xc0
[  555.325794]  [<ffffffff813e96b9>] pci_device_probe+0xf9/0x150
[  555.325799]  [<ffffffff814d7e73>] driver_probe_device+0xa3/0x400
[  555.325802]  [<ffffffff814d82ab>] __driver_attach+0x9b/0xa0
[  555.325805]  [<ffffffff814d8210>] ? __device_attach+0x40/0x40
[  555.325808]  [<ffffffff814d5bd3>] bus_for_each_dev+0x73/0xc0
[  555.325811]  [<ffffffff814d78ce>] driver_attach+0x1e/0x20
[  555.325815]  [<ffffffff814d7480>] bus_add_driver+0x180/0x250
[  555.325819]  [<ffffffffa03b2000>] ? 0xffffffffa03b2000
[  555.325823]  [<ffffffff814d8aa4>] driver_register+0x64/0xf0
[  555.325826]  [<ffffffff813e7bec>] __pci_register_driver+0x4c/0x50
[  555.325832]  [<ffffffffa03d6ca3>] fm10k_register_pci_driver+0x23/0x30 [fm10k]
[  555.325838]  [<ffffffffa03b2080>] fm10k_init_module+0x80/0x1000 [fm10k]
[  555.325843]  [<ffffffff81002128>] do_one_initcall+0xb8/0x200
[  555.325848]  [<ffffffff811e10d2>] ? __vunmap+0xa2/0x100
[  555.325852]  [<ffffffff811fe239>] ? kmem_cache_alloc_trace+0x1b9/0x240
[  555.325855]  [<ffffffff8178230e>] ? do_init_module+0x28/0x1cb
[  555.325858]  [<ffffffff81782346>] do_init_module+0x60/0x1cb
[  555.325862]  [<ffffffff8112168e>] load_module+0x205e/0x26b0
[  555.325866]  [<ffffffff8111d110>] ? store_uevent+0x70/0x70
[  555.325870]  [<ffffffff812234b0>] ? kernel_read+0x50/0x80
[  555.325873]  [<ffffffff81121f3e>] SyS_finit_module+0xbe/0xf0
[  555.325878]  [<ffffffff81789749>] system_call_fastpath+0x12/0x17
[  555.325880] ---[ end trace 9e0f58d071eafd2a ]---

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit e72319bba814b115c47785b3b88f7263d0b8a1b8)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: prevent null pointer dereference of msix_entries table
Jacob Keller [Thu, 4 Feb 2016 18:47:54 +0000 (10:47 -0800)]
fm10k: prevent null pointer dereference of msix_entries table

According to the C standard dereferencing a variable before it is
checked invokes undefined behavior, and thus compilers are free to
assume the check for NULL isn't necessary. Prevent this by re-ordering
the NULL check of msix_entries in fm10k_free_mbx_irq.

Signed-off-by: Jacob Keller <jacob.e.keller@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit de66c610a6adade32bf955f67b4f4f4aaeeeff85)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: use ether_addr_copy to copy MAC address
Bruce Allan [Tue, 29 Dec 2015 02:00:30 +0000 (18:00 -0800)]
fm10k: use ether_addr_copy to copy MAC address

Cleanup the remaining instances of using memcpy() instead of the preferred
ether_addr_copy().

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 11c49f79b294081010f7e13a95c6b40c4d36b1de)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: cleanup SPACE_BEFORE_TAB checkpatch warning
Bruce Allan [Tue, 22 Dec 2015 22:55:26 +0000 (14:55 -0800)]
fm10k: cleanup SPACE_BEFORE_TAB checkpatch warning

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 1905add427cdee1b52380241574ab1339b1df413)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: demote BUG_ON() to WARN_ON() where appropriate
Bruce Allan [Tue, 22 Dec 2015 22:55:20 +0000 (14:55 -0800)]
fm10k: demote BUG_ON() to WARN_ON() where appropriate

We don't need to crash the kernel in this instance so just warn about the
condition and play on.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 838e6102920a288a88f5bba10784ab10b2f2eb3e)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: cleanup remaining right-bit-shifted 1
Bruce Allan [Tue, 22 Dec 2015 21:43:49 +0000 (13:43 -0800)]
fm10k: cleanup remaining right-bit-shifted 1

Use BIT() macro instead.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit fcdb0a9951d8a5edfc47e89a7fe62457c25e18c4)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: Move constants to the right of binary operators
Bruce Allan [Tue, 22 Dec 2015 21:43:44 +0000 (13:43 -0800)]
fm10k: Move constants to the right of binary operators

The semantic patch that makes this change is available
in scripts/coccinelle/misc/compare_const_fl.cocci.

More information about semantic patching is available at
http://coccinelle.lip6.fr/

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 1aab144c507a9849d5b4557d6d78db185ceaef37)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: IS_ENABLED() is not appropriate for boolean kconfig option
Bruce Allan [Wed, 9 Dec 2015 01:20:49 +0000 (17:20 -0800)]
fm10k: IS_ENABLED() is not appropriate for boolean kconfig option

Tri-states need 'if IS_ENABLED()', booleans should use 'ifdef'.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit 0d722ec8bf46cb6547d10e8c5d9b8b6498bc7f97)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: cleanup mailbox code comments etc
Bruce Allan [Wed, 9 Dec 2015 01:20:44 +0000 (17:20 -0800)]
fm10k: cleanup mailbox code comments etc

Cleanup a number of issues with function header comments, lower-case
acronyms (i.e. FIFO, TLV), duplicate comments and a stubbed-out header
comment for fm10k_sm_mbx_init.

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit f632fed30f8e0c1b5c9de209f00145f516e7ad37)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>
8 years agofm10k: use true/false for boolean get_host_state
Bruce Allan [Tue, 8 Dec 2015 23:51:11 +0000 (15:51 -0800)]
fm10k: use true/false for boolean get_host_state

Signed-off-by: Bruce Allan <bruce.w.allan@intel.com>
Tested-by: Krishneil Singh <Krishneil.k.singh@intel.com>
Signed-off-by: Jeff Kirsher <jeffrey.t.kirsher@intel.com>
Orabug: 25394529
(cherry picked from commit f355bb51794af64ef583c259469a778e606d95bb)
Signed-off-by: Jack Vogel <jack.vogel@oracle.com>