linux-2.6-microblaze.git
5 years agohabanalabs: Handle HW_IP_INFO if device disabled or in reset
Tomer Tayar [Thu, 8 Aug 2019 12:30:22 +0000 (12:30 +0000)]
habanalabs: Handle HW_IP_INFO if device disabled or in reset

The HW IP information is relevant even if the device is disabled or in
reset, so always handle the corresponding INFO IOCTL opcode.

Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
5 years agohabanalabs: Expose devices after initialization is done
Tomer Tayar [Thu, 8 Aug 2019 12:25:52 +0000 (12:25 +0000)]
habanalabs: Expose devices after initialization is done

The char devices are currently exposed to user before the device and
driver initialization are done.
This patch moves the cdev and device adding to the system to the end of
the initialization sequence, while keeping the creation of the
structures at the beginning to allow the usage of dev_*().

Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
5 years agohabanalabs: improve security in Debug IOCTL
Omer Shpigelman [Tue, 6 Aug 2019 07:37:59 +0000 (07:37 +0000)]
habanalabs: improve security in Debug IOCTL

This patch improves the security in the Debug IOCTL.
It adds checks that:
- The register index value is in the allowed range for all opcodes.
- The event types number is in the allowed range in SPMU enable.
- The events number is in the allowed range in SPMU disable.

Signed-off-by: Omer Shpigelman <oshpigelman@habana.ai>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
5 years agohabanalabs: use default structure for user input in Debug IOCTL
Omer Shpigelman [Tue, 6 Aug 2019 07:37:59 +0000 (07:37 +0000)]
habanalabs: use default structure for user input in Debug IOCTL

This patch fixes a possible kernel crash when a user provides a too small
input structure to the Debug IOCTL.
The fix sets a default input structure and copies to it the user data.
In case the user provided as input a too small structure, the code will
use the default values taken from the default structure.
Note that in contrary to the input structure, the user can provide an
output structure with changing size or no size at all. Therefore the user
output structure validation is already done in the Debug logic later on.

Signed-off-by: Omer Shpigelman <oshpigelman@habana.ai>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
5 years agohabanalabs: Add descriptive name to PSOC app status register
Tomer Tayar [Thu, 1 Aug 2019 14:28:45 +0000 (14:28 +0000)]
habanalabs: Add descriptive name to PSOC app status register

Add a meaningful name to the general PSOC application status register
which better describes its usage in keeping the HW state.

Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
5 years agohabanalabs: Add descriptive names to PSOC scratch-pad registers
Tomer Tayar [Thu, 1 Aug 2019 14:28:45 +0000 (14:28 +0000)]
habanalabs: Add descriptive names to PSOC scratch-pad registers

The PSOC scratch-pad registers are used for communication with the
device CPU. This patch adds new definitions for these registers which
are more descriptive than their general names.

The new set of definitions also gathers and documents the current usage
of the scratch-pad registers by the driver and the device CPU.

Signed-off-by: Tomer Tayar <ttayar@habana.ai>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
5 years agohabanalabs: create two char devices per ASIC
Oded Gabbay [Tue, 30 Jul 2019 06:10:50 +0000 (09:10 +0300)]
habanalabs: create two char devices per ASIC

This patch changes the driver to create two char devices for each ASIC
it discovers. This is done to allow system/monitoring applications to
query the device for stats, information, idle state and more, while also
allowing the deep-learning application to send work to the ASIC.

One char device is the original device, hlX. IOCTL calls through this
device file can perform any task on the device (compute, memory, queries).
The open function for this device will fail if it was called before but
the file-descriptor it created was not completely released yet (the
release callback function is not called from the kernel until all
instances of that FD are closed). The driver needs to keep this behavior
to support backward compatibility with existing userspace, which count
that the open will fail if the device is "occupied".

The second char device is called "hl_controlDx", where x is the same index
of the main device with a minor number of the original char device + 1.
Applications that open this device can only call the INFO IOCTL. There is
no limitation on the number of applications opening this device.

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agohabanalabs: change device_setup_cdev() to be more generic
Oded Gabbay [Tue, 30 Jul 2019 06:10:02 +0000 (09:10 +0300)]
habanalabs: change device_setup_cdev() to be more generic

This patch re-factors the device_setup_cdev() function to make it more
generic. It doesn't manipulate members of the driver's internal device
structure but instead works only on the arguments that are sent to it.

This is in preparation for using this function to create an additional
char device per ASIC.

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agohabanalabs: maintain a list of file private data objects
Oded Gabbay [Tue, 30 Jul 2019 08:56:09 +0000 (11:56 +0300)]
habanalabs: maintain a list of file private data objects

This patch adds a new list to the driver's device structure. The list will
keep the file private data structures that the driver creates when a user
process opens the device.

This change is needed because it is useless to try to count how many FD
are open. Instead, track our own private data structure per open file and
once it is released, remove it from the list. As long as the list is not
empty, it means we have a user that can do something with our device.

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agohabanalabs: rename user_ctx as compute_ctx
Oded Gabbay [Tue, 30 Jul 2019 08:49:36 +0000 (11:49 +0300)]
habanalabs: rename user_ctx as compute_ctx

This patch renames the "user_ctx" field in the device structure to
"compute_ctx". This better reflects the meaning of this context.

In addition, we also check in the ctx_fini() that the debug mode should be
disabled only if the context being destroyed is the compute context. This
has no effect right now as we only have a single process and a single
context, but this makes the code more ready for multiple process support.

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agohabanalabs: show the process context dram usage
Oded Gabbay [Tue, 30 Jul 2019 08:48:02 +0000 (11:48 +0300)]
habanalabs: show the process context dram usage

When the user query the dram usage of a context, show it the dram usage of
its context, not the user context that is currently running on the device.

This has no effect right now as we only have a single process and a single
context, but this makes the code more ready for multiple process support.

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agohabanalabs: kill user process after CS rollback
Oded Gabbay [Mon, 22 Jul 2019 14:37:22 +0000 (17:37 +0300)]
habanalabs: kill user process after CS rollback

This patch calls the kill user process function after we rollback the
in-flight CSs. This is because the user process can't be closed while
there are open CSs. Therefore, there is no point of sending it a SIGKILL
before we do the rollback CS part.

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agohabanalabs: add handle field to context structure
Oded Gabbay [Mon, 15 Jul 2019 18:55:57 +0000 (21:55 +0300)]
habanalabs: add handle field to context structure

This patch adds a field to the context's structure that will hold a unique
handle for the context.

This will be needed when the user will create the context.

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
Reviewed-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agohabanalabs: Use dev_get_drvdata
Chuhong Yuan [Tue, 23 Jul 2019 12:46:08 +0000 (20:46 +0800)]
habanalabs: Use dev_get_drvdata

Instead of using to_pci_dev + pci_get_drvdata,
use dev_get_drvdata to make code simpler.

Signed-off-by: Chuhong Yuan <hslester96@gmail.com>
Reviewed-by: Oded Gabbay <oded.gabbay@gmail.com>
Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
5 years agohabanalabs: power management through sysfs is only for GOYA
Oded Gabbay [Thu, 4 Jul 2019 08:57:12 +0000 (11:57 +0300)]
habanalabs: power management through sysfs is only for GOYA

The ability of setting power management properties by the system
administrator (through sysfs properties) is only relevant for the GOYA
ASIC. Therefore, move the relevant sysfs properties to the GOYA sysfs
specific file, to make the properties appear in sysfs only for GOYA cards.

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
Reviewed-by: Omer Shpigelman <oshpigelman@habana.ai>
5 years agohabanalabs: cap simulator timeout
Oded Gabbay [Thu, 18 Jul 2019 08:14:37 +0000 (11:14 +0300)]
habanalabs: cap simulator timeout

In the driver timeout functions, we give the simulator a factor of 10
in the timeout. This was necessary when the requested timeout is small
but if it was a few seconds, this can result in a very large timeout which
is unnecessary.

This patch caps the maximum timeout of the simulator to 10 seconds, which
is our largest timeout in the code. That is more then enough for anything
the simulator is doing.

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
Reviewed-by: Omer Shpigelman <oshpigelman@habana.ai>
5 years agohabanalabs: add debug print when rejecting CS
Oded Gabbay [Thu, 18 Jul 2019 06:03:58 +0000 (09:03 +0300)]
habanalabs: add debug print when rejecting CS

When rejecting CS because of too many in-flight CS, print a debug message
about it as it useful to know when the user is debugging (it indicates a
back-pressure from the driver as the device is not fast enough to consume
the CS)

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
Reviewed-by: Omer Shpigelman <oshpigelman@habana.ai>
5 years agohabanalabs: add comments on INFO IOCTL
Oded Gabbay [Tue, 16 Jul 2019 05:55:04 +0000 (08:55 +0300)]
habanalabs: add comments on INFO IOCTL

This patch adds some in-code documentation on the different opcodes of the
INFO IOCTL.

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
Reviewed-by: Omer Shpigelman <oshpigelman@habana.ai>
5 years agohabanalabs: remove write_open_cnt property
Oded Gabbay [Mon, 8 Jul 2019 11:53:43 +0000 (14:53 +0300)]
habanalabs: remove write_open_cnt property

This property has attempted to show the number of open file descriptors on
the device. This was a stupid and futile attempt so remove this property
completely.

Signed-off-by: Oded Gabbay <oded.gabbay@gmail.com>
5 years agow1: add DS2501, DS2502, DS2505 EPROM device driver
Thomas Bogendoerfer [Sat, 31 Aug 2019 08:26:22 +0000 (10:26 +0200)]
w1: add DS2501, DS2502, DS2505 EPROM device driver

Add a 1-Wire slave driver to support DS250x EPROM deivces. This
slave driver attaches the devices to the NVMEM subsystem for
an easy in-kernel usage.

Signed-off-by: Thomas Bogendoerfer <tbogendoerfer@suse.de>
Link: https://lore.kernel.org/r/20190831082623.15627-3-tbogendoerfer@suse.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agow1: add 1-wire master driver for IP block found in SGI ASICs
Thomas Bogendoerfer [Sat, 31 Aug 2019 08:26:21 +0000 (10:26 +0200)]
w1: add 1-wire master driver for IP block found in SGI ASICs

Starting with SGI Origin machines nearly every new SGI ASIC contains
an 1-Wire master. They are used for attaching One-Wire prom devices,
which contain information about part numbers, revision numbers,
serial number etc. and MAC addresses for ethernet interfaces.
This patch adds a master driver to support this IP block.
It also adds an extra field dev_id to struct w1_bus_master, which
could be in used in slave drivers for creating unique device names.

Signed-off-by: Thomas Bogendoerfer <tbogendoerfer@suse.de>
Link: https://lore.kernel.org/r/20190831082623.15627-2-tbogendoerfer@suse.de
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agopcmcia/i82092: Refactored dprintk macro for dev_dbg().
Adam Zerella [Sun, 25 Aug 2019 05:35:10 +0000 (15:35 +1000)]
pcmcia/i82092: Refactored dprintk macro for dev_dbg().

As suggested in https://kernelnewbies.org/KernelJanitors/Todo
this patch replaces the outdated macro of DPRINTK for dev_dbg()

To: Dominik Brodowski <linux@dominikbrodowski.net>
To: Thomas Gleixner <tglx@linutronix.de>
To: Adam Zerella <adam.zerella@gmail.com>
To: linux-kernel@vger.kernel.org
Signed-off-by: Adam Zerella <adam.zerella@gmail.com>
Link: https://lore.kernel.org/r/20190825053513.13990-1-adam.zerella@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agouio: Documentation: Add information on using uio_pdrv_genirq with DT
Daniel Mack [Thu, 15 Aug 2019 21:28:07 +0000 (23:28 +0200)]
uio: Documentation: Add information on using uio_pdrv_genirq with DT

Add a paragraph to describe the use of the "of_id" module parameter,
along with the new DT property.

Signed-off-by: Daniel Mack <daniel@zonque.org>
Link: https://lore.kernel.org/r/20190815212807.25058-2-daniel@zonque.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agouio: uio_pdrv_genirq: Make UIO name controllable via DT node property
Daniel Mack [Thu, 15 Aug 2019 21:28:06 +0000 (23:28 +0200)]
uio: uio_pdrv_genirq: Make UIO name controllable via DT node property

When probed via DT, the uio_pdrv_genirq driver currently uses the name
of the node and exposes that as name of the UIO device to userspace.

This doesn't work for systems where multiple nodes with the same name
(but different unit addresses) are present, or for systems where the
node names are auto-generated by a third-party tool.

This patch adds the possibility to read the UIO name from the optional
"linux,uio-name" property.

Signed-off-by: Daniel Mack <daniel@zonque.org>
Link: https://lore.kernel.org/r/20190815212807.25058-1-daniel@zonque.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agotoshiba: Add correct printk log level while emitting error log
Rishi Gupta [Sun, 18 Aug 2019 07:34:31 +0000 (13:04 +0530)]
toshiba: Add correct printk log level while emitting error log

The printk functions are invoked without specifying required
log level when printing error messages. This commit replaces
all direct uses of printk with their corresponding pr_err/info/debug
variant.

Signed-off-by: Rishi Gupta <gupt21@gmail.com>
Link: https://lore.kernel.org/r/1566113671-8743-1-git-send-email-gupt21@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agomisc: fastrpc: free dma buf scatter list
Srinivas Kandagatla [Thu, 29 Aug 2019 09:29:26 +0000 (10:29 +0100)]
misc: fastrpc: free dma buf scatter list

dma buf scatter list is never freed, free it!

Orignally detected by kmemleak:
  backtrace:
    [<ffffff80088b7658>] kmemleak_alloc+0x50/0x84
    [<ffffff8008373284>] sg_kmalloc+0x38/0x60
    [<ffffff8008373144>] __sg_alloc_table+0x60/0x110
    [<ffffff800837321c>] sg_alloc_table+0x28/0x58
    [<ffffff800837336c>] __sg_alloc_table_from_pages+0xc0/0x1ac
    [<ffffff800837346c>] sg_alloc_table_from_pages+0x14/0x1c
    [<ffffff8008097a3c>] __iommu_get_sgtable+0x5c/0x8c
    [<ffffff800850a1d0>] fastrpc_dma_buf_attach+0x84/0xf8
    [<ffffff80085114bc>] dma_buf_attach+0x70/0xc8
    [<ffffff8008509efc>] fastrpc_map_create+0xf8/0x1e8
    [<ffffff80085086f4>] fastrpc_device_ioctl+0x508/0x900
    [<ffffff80082428c8>] compat_SyS_ioctl+0x128/0x200
    [<ffffff80080832c4>] el0_svc_naked+0x34/0x38
    [<ffffffffffffffff>] 0xffffffffffffffff

Reported-by: Mayank Chopra <mak.chopra@codeaurora.org>
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Link: https://lore.kernel.org/r/20190829092926.12037-6-srinivas.kandagatla@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agomisc: fastrpc: fix double refcounting on dmabuf
Srinivas Kandagatla [Thu, 29 Aug 2019 09:29:25 +0000 (10:29 +0100)]
misc: fastrpc: fix double refcounting on dmabuf

dma buf refcount has to be done by the driver which is going to use the fd.
This driver already does refcount on the dmabuf fd if its actively using it
but also does an additional refcounting via extra ioctl.
This additional refcount can lead to memory leak in cases where the
applications fail to call the ioctl to decrement the refcount.

So remove this extra refcount in the ioctl

More info of dma buf usage at drivers/dma-buf/dma-buf.c

Reported-by: Mayank Chopra <mak.chopra@codeaurora.org>
Reported-by: Jorge Ramirez-Ortiz <jorge.ramirez-ortiz@linaro.org>
Tested-by: Jorge Ramirez-Ortiz <jorge.ramirez-ortiz@linaro.org>
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Link: https://lore.kernel.org/r/20190829092926.12037-5-srinivas.kandagatla@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agomisc: fastrpc: remove unused definition
Jorge Ramirez-Ortiz [Thu, 29 Aug 2019 09:29:24 +0000 (10:29 +0100)]
misc: fastrpc: remove unused definition

Remove unused INIT_MEMLEN_MAX define.

Signed-off-by: Jorge Ramirez-Ortiz <jorge.ramirez-ortiz@linaro.org>
Signed-off-by: Abhinav Asati <asatiabhi@codeaurora.org>
Signed-off-by: Vamsi Singamsetty <vamssi@codeaurora.org>
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Link: https://lore.kernel.org/r/20190829092926.12037-4-srinivas.kandagatla@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agomisc: fastrpc: Don't reference rpmsg_device after remove
Bjorn Andersson [Thu, 29 Aug 2019 09:29:23 +0000 (10:29 +0100)]
misc: fastrpc: Don't reference rpmsg_device after remove

As fastrpc_rpmsg_remove() returns the rpdev of the channel context is no
longer a valid object, so ensure to update the channel context to no
longer reference the old object and guard in the invoke code path
against dereferencing it.

Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Mayank Chopra <mak.chopra@codeaurora.org>
Signed-off-by: Abhinav Asati <asatiabhi@codeaurora.org>
Signed-off-by: Vamsi Singamsetty <vamssi@codeaurora.org>
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Link: https://lore.kernel.org/r/20190829092926.12037-3-srinivas.kandagatla@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agomisc: fastrpc: Reference count channel context
Bjorn Andersson [Thu, 29 Aug 2019 09:29:22 +0000 (10:29 +0100)]
misc: fastrpc: Reference count channel context

The channel context is referenced from the fastrpc user and might as
user space holds the file descriptor open outlive the fastrpc device,
which is removed when the remote processor is shutting down.

Reference count the channel context in order to retain this object until
all references has been relinquished.

Signed-off-by: Bjorn Andersson <bjorn.andersson@linaro.org>
Signed-off-by: Mayank Chopra <mak.chopra@codeaurora.org>
Signed-off-by: Abhinav Asati <asatiabhi@codeaurora.org>
Signed-off-by: Vamsi Singamsetty <vamssi@codeaurora.org>
Signed-off-by: Srinivas Kandagatla <srinivas.kandagatla@linaro.org>
Link: https://lore.kernel.org/r/20190829092926.12037-2-srinivas.kandagatla@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agoMAINTAINERS: add maintainer for Intel Stratix10 FW drivers
Richard Gong [Tue, 3 Sep 2019 13:18:21 +0000 (08:18 -0500)]
MAINTAINERS: add maintainer for Intel Stratix10 FW drivers

Add myself as maintainer for the newly created Intel Stratix10
firmware drivers.

Signed-off-by: Richard Gong <richard.gong@intel.com>
Reviewed-by: Alan Tull <atull@kernel.org>
Link: https://lore.kernel.org/r/1567516701-26026-5-git-send-email-richard.gong@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agofirmware: rsu: document sysfs interface
Richard Gong [Tue, 3 Sep 2019 13:18:20 +0000 (08:18 -0500)]
firmware: rsu: document sysfs interface

Describe Intel Stratix10 Remote System Update (RSU) device attributes

Signed-off-by: Richard Gong <richard.gong@intel.com>
Reviewed-by: Alan Tull <atull@kernel.org>
Link: https://lore.kernel.org/r/1567516701-26026-4-git-send-email-richard.gong@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agofirmware: add Intel Stratix10 remote system update driver
Richard Gong [Tue, 3 Sep 2019 13:18:19 +0000 (08:18 -0500)]
firmware: add Intel Stratix10 remote system update driver

The Intel Remote System Update (RSU) driver exposes interfaces access
through the Intel Service Layer to user space via sysfs interface.
The RSU interfaces report and control some of the optional RSU features
on Intel Stratix 10 SoC.

The RSU feature provides a way for customers to update the boot
configuration of a Intel Stratix 10 SoC device with significantly reduced
risk of corrupting the bitstream storage and bricking the system.

Signed-off-by: Richard Gong <richard.gong@intel.com>
Reviewed-by: Alan Tull <atull@kernel.org>
Link: https://lore.kernel.org/r/1567516701-26026-3-git-send-email-richard.gong@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agofirmware: stratix10-svc: extend svc to support new RSU features
Richard Gong [Tue, 3 Sep 2019 13:18:18 +0000 (08:18 -0500)]
firmware: stratix10-svc: extend svc to support new RSU features

Extend Intel Stratix10 service layer driver to support new RSU notify and
MAX_RETRY with watchdog event.

RSU is used to provide our customers with protection against loading bad
bitstream onto their devices when those devices are booting from flash

RSU notifies provides users with an API to notify the firmware of the
state of hard processor system.

To deal with watchdog event, RSU provides a way for user to retry the
current running image several times before giving up and starting normal
RSU failover flow.

Signed-off-by: Richard Gong <richard.gong@intel.com>
Reviewed-by: Alan Tull <atull@kernel.org>
Link: https://lore.kernel.org/r/1567516701-26026-2-git-send-email-richard.gong@linux.intel.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agoselftests: firmware: Add request_firmware_into_buf tests
Scott Branden [Thu, 22 Aug 2019 18:40:05 +0000 (11:40 -0700)]
selftests: firmware: Add request_firmware_into_buf tests

Add tests cases for checking request_firmware_into_buf api.
API was introduced into kernel with no testing present previously.

Signed-off-by: Scott Branden <scott.branden@broadcom.com>
Acked-by: Luis Chamberlain <mcgrof@kernel.org>
Acked-by: Shuah Khan <skhan@linuxfoundation.org>
Link: https://lore.kernel.org/r/20190822184005.901-3-scott.branden@broadcom.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agotest_firmware: add support for request_firmware_into_buf
Scott Branden [Thu, 22 Aug 2019 18:40:04 +0000 (11:40 -0700)]
test_firmware: add support for request_firmware_into_buf

Add test config into_buf to allow request_firmware_into_buf to be
called instead of request_firmware/request_firmware_direct.  The number
of parameters differ calling request_firmware_into_buf and support
has not been added to test such api in test_firmware until now.

Signed-off-by: Scott Branden <scott.branden@broadcom.com>
Acked-by: Luis Chamberlain <mcgrof@kernel.org>
Acked-by: Shuah Khan <skhan@linuxfoundation.org>
Link: https://lore.kernel.org/r/20190822184005.901-2-scott.branden@broadcom.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agofirmware: google: check if size is valid when decoding VPD data
Hung-Te Lin [Fri, 30 Aug 2019 02:23:58 +0000 (10:23 +0800)]
firmware: google: check if size is valid when decoding VPD data

The VPD implementation from Chromium Vital Product Data project used to
parse data from untrusted input without checking if the meta data is
invalid or corrupted. For example, the size from decoded content may
be negative value, or larger than whole input buffer. Such invalid data
may cause buffer overflow.

To fix that, the size parameters passed to vpd_decode functions should
be changed to unsigned integer (u32) type, and the parsing of entry
header should be refactored so every size field is correctly verified
before starting to decode.

Fixes: ad2ac9d5c5e0 ("firmware: Google VPD: import lib_vpd source files")
Signed-off-by: Hung-Te Lin <hungte@chromium.org>
Cc: stable <stable@vger.kernel.org>
Reviewed-by: Guenter Roeck <linux@roeck-us.net>
Reviewed-by: Stephen Boyd <swboyd@chromium.org>
Link: https://lore.kernel.org/r/20190830022402.214442-1-hungte@chromium.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years ago/dev/mem: Bail out upon SIGKILL.
Tetsuo Handa [Mon, 26 Aug 2019 13:13:25 +0000 (22:13 +0900)]
/dev/mem: Bail out upon SIGKILL.

syzbot found that a thread can stall for minutes inside read_mem() or
write_mem() after that thread was killed by SIGKILL [1]. Reading from
iomem areas of /dev/mem can be slow, depending on the hardware.
While reading 2GB at one read() is legal, delaying termination of killed
thread for minutes is bad. Thus, allow reading/writing /dev/mem and
/dev/kmem to be preemptible and killable.

  [ 1335.912419][T20577] read_mem: sz=4096 count=2134565632
  [ 1335.943194][T20577] read_mem: sz=4096 count=2134561536
  [ 1335.978280][T20577] read_mem: sz=4096 count=2134557440
  [ 1336.011147][T20577] read_mem: sz=4096 count=2134553344
  [ 1336.041897][T20577] read_mem: sz=4096 count=2134549248

Theoretically, reading/writing /dev/mem and /dev/kmem can become
"interruptible". But this patch chose "killable". Future patch will make
them "interruptible" so that we can revert to "killable" if some program
regressed.

[1] https://syzkaller.appspot.com/bug?id=a0e3436829698d5824231251fad9d8e998f94f5e

Signed-off-by: Tetsuo Handa <penguin-kernel@I-love.SAKURA.ne.jp>
Cc: stable <stable@vger.kernel.org>
Reported-by: syzbot <syzbot+8ab2d0f39fb79fe6ca40@syzkaller.appspotmail.com>
Link: https://lore.kernel.org/r/1566825205-10703-1-git-send-email-penguin-kernel@I-love.SAKURA.ne.jp
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agobinder: Add binder_proc logging to binderfs
Hridya Valsaraju [Tue, 3 Sep 2019 16:16:55 +0000 (09:16 -0700)]
binder: Add binder_proc logging to binderfs

Currently /sys/kernel/debug/binder/proc contains
the debug data for every binder_proc instance.
This patch makes this information also available
in a binderfs instance mounted with a mount option
"stats=global" in addition to debugfs. The patch does
not affect the presence of the file in debugfs.

If a binderfs instance is mounted at path /dev/binderfs,
this file would be present at /dev/binderfs/binder_logs/proc.
This change provides an alternate way to access this file when debugfs
is not mounted.

Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
Signed-off-by: Hridya Valsaraju <hridya@google.com>
Link: https://lore.kernel.org/r/20190903161655.107408-5-hridya@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agobinder: Make transaction_log available in binderfs
Hridya Valsaraju [Tue, 3 Sep 2019 16:16:54 +0000 (09:16 -0700)]
binder: Make transaction_log available in binderfs

Currently, the binder transaction log files 'transaction_log'
and 'failed_transaction_log' live in debugfs at the following locations:

/sys/kernel/debug/binder/failed_transaction_log
/sys/kernel/debug/binder/transaction_log

This patch makes these files also available in a binderfs instance
mounted with the mount option "stats=global".
It does not affect the presence of these files in debugfs.
If a binderfs instance is mounted at path /dev/binderfs, the location of
these files will be as follows:

/dev/binderfs/binder_logs/failed_transaction_log
/dev/binderfs/binder_logs/transaction_log

This change provides an alternate option to access these files when
debugfs is not mounted.

Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
Signed-off-by: Hridya Valsaraju <hridya@google.com>
Link: https://lore.kernel.org/r/20190903161655.107408-4-hridya@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agobinder: Add stats, state and transactions files
Hridya Valsaraju [Tue, 3 Sep 2019 16:16:53 +0000 (09:16 -0700)]
binder: Add stats, state and transactions files

The following binder stat files currently live in debugfs.

/sys/kernel/debug/binder/state
/sys/kernel/debug/binder/stats
/sys/kernel/debug/binder/transactions

This patch makes these files available in a binderfs instance
mounted with the mount option 'stats=global'. For example, if a binderfs
instance is mounted at path /dev/binderfs, the above files will be
available at the following locations:

/dev/binderfs/binder_logs/state
/dev/binderfs/binder_logs/stats
/dev/binderfs/binder_logs/transactions

This provides a way to access them even when debugfs is not mounted.

Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
Signed-off-by: Hridya Valsaraju <hridya@google.com>
Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
Link: https://lore.kernel.org/r/20190903161655.107408-3-hridya@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agobinder: add a mount option to show global stats
Hridya Valsaraju [Tue, 3 Sep 2019 16:16:52 +0000 (09:16 -0700)]
binder: add a mount option to show global stats

Currently, all binder state and statistics live in debugfs.
We need this information even when debugfs is not mounted.
This patch adds the mount option 'stats' to enable a binderfs
instance to have binder debug information present in the same.
'stats=global' will enable the global binder statistics. In
the future, 'stats=local' will enable binder statistics local
to the binderfs instance. The two modes 'global' and 'local'
will be mutually exclusive. 'stats=global' option is only available
for a binderfs instance mounted in the initial user namespace.
An attempt to use the option to mount a binderfs instance in
another user namespace will return an EPERM error.

Signed-off-by: Hridya Valsaraju <hridya@google.com>
Acked-by: Christian Brauner <christian.brauner@ubuntu.com>
Link: https://lore.kernel.org/r/20190903161655.107408-2-hridya@google.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agobinder: Add default binder devices through binderfs when configured
Hridya Valsaraju [Wed, 4 Sep 2019 11:07:03 +0000 (13:07 +0200)]
binder: Add default binder devices through binderfs when configured

Currently, since each binderfs instance needs its own
private binder devices, every time a binderfs instance is
mounted, all the default binder devices need to be created
via the BINDER_CTL_ADD IOCTL. This patch aims to
add a solution to automatically create the default binder
devices for each binderfs instance that gets mounted.
To achieve this goal, when CONFIG_ANDROID_BINDERFS is set,
the default binder devices specified by CONFIG_ANDROID_BINDER_DEVICES
are created in each binderfs instance instead of global devices
being created by the binder driver.

Co-developed-by: Christian Brauner <christian.brauner@ubuntu.com>
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Signed-off-by: Hridya Valsaraju <hridya@google.com>
Reviewed-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Link: https://lore.kernel.org/r/20190808222727.132744-2-hridya@google.com
Link: https://lore.kernel.org/r/20190904110704.8606-2-christian.brauner@ubuntu.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agobinder: Validate the default binderfs device names.
Hridya Valsaraju [Wed, 4 Sep 2019 11:07:04 +0000 (13:07 +0200)]
binder: Validate the default binderfs device names.

Length of a binderfs device name cannot exceed BINDERFS_MAX_NAME.
This patch adds a check in binderfs_init() to ensure the same
for the default binder devices that will be created in every
binderfs instance.

Co-developed-by: Christian Brauner <christian.brauner@ubuntu.com>
Signed-off-by: Christian Brauner <christian.brauner@ubuntu.com>
Signed-off-by: Hridya Valsaraju <hridya@google.com>
Reviewed-by: Joel Fernandes (Google) <joel@joelfernandes.org>
Link: https://lore.kernel.org/r/20190808222727.132744-3-hridya@google.com
Link: https://lore.kernel.org/r/20190904110704.8606-3-christian.brauner@ubuntu.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agoeeprom: Deprecate the legacy eeprom driver
Jean Delvare [Mon, 2 Sep 2019 08:48:38 +0000 (10:48 +0200)]
eeprom: Deprecate the legacy eeprom driver

Time has come to get rid of the old eeprom driver. The at24 driver
should be used instead. So mark the eeprom driver as deprecated and
give users some time to migrate. Then we can remove the legacy
eeprom driver completely.

Signed-off-by: Jean Delvare <jdelvare@suse.de>
Cc: Arnd Bergmann <arnd@arndb.de>
Link: https://lore.kernel.org/r/20190902104838.058725c2@endymion
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agoMerge tag 'fpga-dfl-for-5.4' of git://git.kernel.org/pub/scm/linux/kernel/git/mdf...
Greg Kroah-Hartman [Wed, 4 Sep 2019 05:43:42 +0000 (07:43 +0200)]
Merge tag 'fpga-dfl-for-5.4' of git://git./linux/kernel/git/mdf/linux-fpga into char-misc-next

Moritz writes:

FPGA DFL Changes for 5.4

This pull-request contains the FPGA DFL changes for 5.4

- The first three patches are cleanup patches making use of dev_groups and
  making the init callback optional.
- One patch adds userclock sysfs entries that are DFL specific
- One patch exposes AFU port disable/enable functions
- One patch adds error reporting
- One patch adds AFU SignalTap support
- One patch adds FME global error reporting
- The final patch is a documentation patch that decribes the
  virtualization interfaces

This patchset requires the 'dev_groups_all_drivers' tag from drivers
core for the dev_groups refactoring as well as the DFL changes already
in char-misc-next.

Signed-off-by: Moritz Fischer <mdf@kernel.org>
* tag 'fpga-dfl-for-5.4' of git://git.kernel.org/pub/scm/linux/kernel/git/mdf/linux-fpga:
  Documentation: fpga: dfl: add descriptions for virtualization and new interfaces.
  fpga: dfl: fme: add global error reporting support
  fpga: dfl: afu: add STP (SignalTap) support
  fpga: dfl: afu: add error reporting support.
  fpga: dfl: afu: expose __afu_port_enable/disable function.
  fpga: dfl: afu: add userclock sysfs interfaces.
  fpga: dfl: afu: convert platform_driver to use dev_groups
  fpga: dfl: fme: convert platform_driver to use dev_groups
  fpga: dfl: make init callback optional
  driver core: add dev_groups to all drivers

5 years agoDocumentation: fpga: dfl: add descriptions for virtualization and new interfaces.
Wu Hao [Mon, 12 Aug 2019 02:50:04 +0000 (10:50 +0800)]
Documentation: fpga: dfl: add descriptions for virtualization and new interfaces.

This patch adds virtualization support description for DFL based
FPGA devices (based on PCIe SRIOV), and introductions to new
interfaces added by new dfl private feature drivers.

[mdf@kernel.org: Fixed up to make it work with new reStructuredText docs]
Signed-off-by: Xu Yilun <yilun.xu@intel.com>
Signed-off-by: Wu Hao <hao.wu@intel.com>
Acked-by: Alan Tull <atull@kernel.org>
Signed-off-by: Moritz Fischer <mdf@kernel.org>
5 years agofpga: dfl: fme: add global error reporting support
Wu Hao [Mon, 12 Aug 2019 02:50:03 +0000 (10:50 +0800)]
fpga: dfl: fme: add global error reporting support

This patch adds support for global error reporting for FPGA
Management Engine (FME), it introduces sysfs interfaces to
report different error detected by the hardware, and allow
user to clear errors or inject error for testing purpose.

Signed-off-by: Luwei Kang <luwei.kang@intel.com>
Signed-off-by: Ananda Ravuri <ananda.ravuri@intel.com>
Signed-off-by: Xu Yilun <yilun.xu@intel.com>
Signed-off-by: Wu Hao <hao.wu@intel.com>
Acked-by: Alan Tull <atull@kernel.org>
Signed-off-by: Moritz Fischer <mdf@kernel.org>
5 years agofpga: dfl: afu: add STP (SignalTap) support
Wu Hao [Mon, 12 Aug 2019 02:50:02 +0000 (10:50 +0800)]
fpga: dfl: afu: add STP (SignalTap) support

STP (SignalTap) is one of the private features under the port for
debugging. This patch adds private feature driver support for it
to allow userspace applications to mmap related mmio region and
provide STP service.

Signed-off-by: Xu Yilun <yilun.xu@intel.com>
Signed-off-by: Wu Hao <hao.wu@intel.com>
Acked-by: Moritz Fischer <mdf@kernel.org>
Acked-by: Alan Tull <atull@kernel.org>
Signed-off-by: Moritz Fischer <mdf@kernel.org>
5 years agofpga: dfl: afu: add error reporting support.
Wu Hao [Mon, 12 Aug 2019 02:50:01 +0000 (10:50 +0800)]
fpga: dfl: afu: add error reporting support.

Error reporting is one important private feature, it reports error
detected on port and accelerated function unit (AFU). It introduces
several sysfs interfaces to allow userspace to check and clear
errors detected by hardware.

Signed-off-by: Xu Yilun <yilun.xu@intel.com>
Signed-off-by: Wu Hao <hao.wu@intel.com>
Acked-by: Alan Tull <atull@kernel.org>
Signed-off-by: Moritz Fischer <mdf@kernel.org>
5 years agofpga: dfl: afu: expose __afu_port_enable/disable function.
Wu Hao [Mon, 12 Aug 2019 02:50:00 +0000 (10:50 +0800)]
fpga: dfl: afu: expose __afu_port_enable/disable function.

As these two functions are used by other private features within the
same driver module but different driver files. e.g. in error reporting
private feature, it requires to clear errors when port is in reset.

Signed-off-by: Xu Yilun <yilun.xu@intel.com>
Signed-off-by: Wu Hao <hao.wu@intel.com>
Acked-by: Moritz Fischer <mdf@kernel.org>
Acked-by: Alan Tull <atull@kernel.org>
Signed-off-by: Moritz Fischer <mdf@kernel.org>
5 years agofpga: dfl: afu: add userclock sysfs interfaces.
Wu Hao [Mon, 12 Aug 2019 02:49:59 +0000 (10:49 +0800)]
fpga: dfl: afu: add userclock sysfs interfaces.

This patch introduces userclock sysfs interfaces for AFU, user
could use these interfaces for clock setting to AFU.

Please note that, this is only working for port header feature
with revision 0, for later revisions, userclock setting is moved
to a separated private feature, so one revision sysfs interface
is exposed to userspace application for this purpose too.

Signed-off-by: Ananda Ravuri <ananda.ravuri@intel.com>
Signed-off-by: Russ Weight <russell.h.weight@intel.com>
Signed-off-by: Xu Yilun <yilun.xu@intel.com>
Signed-off-by: Wu Hao <hao.wu@intel.com>
Acked-by: Alan Tull <atull@kernel.org>
Signed-off-by: Moritz Fischer <mdf@kernel.org>
5 years agofpga: dfl: afu: convert platform_driver to use dev_groups
Wu Hao [Mon, 12 Aug 2019 02:49:58 +0000 (10:49 +0800)]
fpga: dfl: afu: convert platform_driver to use dev_groups

This patch takes advantage of driver core which helps to create
and remove sysfs attribute files, so there is no need to register
sysfs entries manually in dfl-afu platform river code.

Signed-off-by: Wu Hao <hao.wu@intel.com>
Acked-by: Moritz Fischer <mdf@kernel.org>
Signed-off-by: Moritz Fischer <mdf@kernel.org>
5 years agofpga: dfl: fme: convert platform_driver to use dev_groups
Wu Hao [Mon, 12 Aug 2019 02:49:57 +0000 (10:49 +0800)]
fpga: dfl: fme: convert platform_driver to use dev_groups

This patch takes advantage of driver core which helps to create
and remove sysfs attribute files, so there is no need to register
sysfs entries manually in dfl-fme platform river code.

Signed-off-by: Wu Hao <hao.wu@intel.com>
Acked-by: Moritz Fischer <mdf@kernel.org>
Signed-off-by: Moritz Fischer <mdf@kernel.org>
5 years agofpga: dfl: make init callback optional
Wu Hao [Mon, 12 Aug 2019 02:49:56 +0000 (10:49 +0800)]
fpga: dfl: make init callback optional

This patch makes init callback of sub features optional. With
this change, people don't need to prepare any empty init callback.

Signed-off-by: Wu Hao <hao.wu@intel.com>
Acked-by: Moritz Fischer <mdf@kernel.org>
Signed-off-by: Moritz Fischer <mdf@kernel.org>
5 years agoMerge branch 'char-misc-next' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh...
Moritz Fischer [Wed, 4 Sep 2019 02:35:07 +0000 (19:35 -0700)]
Merge branch 'char-misc-next' of git://git./linux/kernel/git/gregkh/char-misc into fpga-dfl-for-5.4

5 years agoMerge tag 'dev_groups_all_drivers' into fpga-dfl-for-5.4
Moritz Fischer [Wed, 4 Sep 2019 02:08:35 +0000 (19:08 -0700)]
Merge tag 'dev_groups_all_drivers' into fpga-dfl-for-5.4

dev_groups added to struct driver

Persistent tag for others to pull this branch from

This is the first patch in a longer series that adds the ability for the
driver core to create and remove a list of attribute groups
automatically when the device is bound/unbound from a specific driver.

See:
https://lore.kernel.org/r/20190731124349.4474-2-gregkh@linuxfoundation.org
for details on this patch, and examples of how to use it in other
drivers.

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agocoresight: tmc-etr: Add barrier packets when moving offset forward
Mathieu Poirier [Thu, 29 Aug 2019 20:28:42 +0000 (14:28 -0600)]
coresight: tmc-etr: Add barrier packets when moving offset forward

This patch adds barrier packets in the trace stream when the offset in the
data buffer needs to be moved forward.  Otherwise the decoder isn't aware
of the break in the stream and can't synchronise itself with the trace
data.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Tested-by: Yabin Cui <yabinc@google.com>
Reviewed-by: Leo Yan <leo.yan@linaro.org>
Link: https://lore.kernel.org/r/20190829202842.580-18-mathieu.poirier@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agocoresight: tmc-etr: Decouple buffer sync and barrier packet insertion
Mathieu Poirier [Thu, 29 Aug 2019 20:28:41 +0000 (14:28 -0600)]
coresight: tmc-etr: Decouple buffer sync and barrier packet insertion

If less space is available in the perf ring buffer than the ETR buffer,
barrier packets inserted in the trace stream by tmc_sync_etr_buf() are
skipped over when the head of the buffer is moved forward, resulting in
traces that can't be decoded.

This patch decouples the process of syncing ETR buffers and the addition
of barrier packets in order to perform the latter once the offset in the
trace buffer has been properly computed.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Reviewed-by: Leo Yan <leo.yan@linaro.org>
Link: https://lore.kernel.org/r/20190829202842.580-17-mathieu.poirier@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agocoresight: tmc: Make memory width mask computation into a function
Mathieu Poirier [Thu, 29 Aug 2019 20:28:40 +0000 (14:28 -0600)]
coresight: tmc: Make memory width mask computation into a function

Make the computation of a memory mask representing the width of the memory
bus into a function so that it can be re-used by the ETR driver.

Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Reviewed-by: Leo Yan <leo.yan@linaro.org>
Link: https://lore.kernel.org/r/20190829202842.580-16-mathieu.poirier@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agocoresight: tmc-etr: Fix perf_data check
Yabin Cui [Thu, 29 Aug 2019 20:28:39 +0000 (14:28 -0600)]
coresight: tmc-etr: Fix perf_data check

When tracing etm data of multiple threads on multiple cpus through
perf interface, each cpu has a unique etr_perf_buffer while sharing
the same etr device. There is no guarantee that the last cpu starts
etm tracing also stops last. This makes perf_data check fail.

Fix it by checking etr_buf instead of etr_perf_buffer.
Also move the code setting and clearing perf_buf to more suitable
places.

Fixes: 3147da92a8a8 ("coresight: tmc-etr: Allocate and free ETR memory buffers for CPU-wide scenarios")
Signed-off-by: Yabin Cui <yabinc@google.com>
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Link: https://lore.kernel.org/r/20190829202842.580-15-mathieu.poirier@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agocoresight: tmc-etr: Fix updating buffer in not-snapshot mode.
Yabin Cui [Thu, 29 Aug 2019 20:28:38 +0000 (14:28 -0600)]
coresight: tmc-etr: Fix updating buffer in not-snapshot mode.

TMC etr always copies all available data to perf aux buffer, which
may exceed the available space in perf aux buffer. It isn't suitable
for not-snapshot mode, because:
1) It may overwrite previously written data.
2) It may make the perf_event_mmap_page->aux_head report having more
or less data than the reality.

So change to only copy the latest data fitting the available space in
perf aux buffer.

Signed-off-by: Yabin Cui <yabinc@google.com>
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Link: https://lore.kernel.org/r/20190829202842.580-14-mathieu.poirier@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agocoresight: etm4x: improve clarity of etm4_os_unlock comment
Andrew Murray [Thu, 29 Aug 2019 20:28:37 +0000 (14:28 -0600)]
coresight: etm4x: improve clarity of etm4_os_unlock comment

To improve clarity, let's update the comment for etm4_os_unlock
to use the name of the register as per the ETM architecture
specification.

The existing comment is also misleading as it suggests any value
written to TRCOSLAR unlocks the trace registers, however it must
be '0' - let's also correct this.

Signed-off-by: Andrew Murray <andrew.murray@arm.com>
Reviewed-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Link: https://lore.kernel.org/r/20190829202842.580-13-mathieu.poirier@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agocoresight: etm4x: use module_param instead of module_param_named
Andrew Murray [Thu, 29 Aug 2019 20:28:36 +0000 (14:28 -0600)]
coresight: etm4x: use module_param instead of module_param_named

Given that the user-exposed module parameter for 'boot_enable' matches
the variable that it sets, let's use module_param instead of
module_param_named.

Let's also use octal permissions (checkpatch recommends this) and
provide a module parameter description.

Signed-off-by: Andrew Murray <andrew.murray@arm.com>
Reviewed-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Link: https://lore.kernel.org/r/20190829202842.580-12-mathieu.poirier@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agocoresight: etm4x: Use explicit barriers on enable/disable
Andrew Murray [Thu, 29 Aug 2019 20:28:35 +0000 (14:28 -0600)]
coresight: etm4x: Use explicit barriers on enable/disable

Synchronization is recommended before disabling the trace registers
to prevent any start or stop points being speculative at the point
of disabling the unit (section 7.3.77 of ARM IHI 0064D).

Synchronization is also recommended after programming the trace
registers to ensure all updates are committed prior to normal code
resuming (section 4.3.7 of ARM IHI 0064D).

Let's ensure these syncronization points are present in the code
and clearly commented.

Note that we could rely on the barriers in CS_LOCK and
coresight_disclaim_device_unlocked or the context switch to user
space - however coresight may be of use in the kernel.

On armv8 the mb macro is defined as dsb(sy) - Given that the etm4x is
only used on armv8 let's directly use dsb(sy) instead of mb(). This
removes some ambiguity and makes it easier to correlate the code with
the TRM.

Signed-off-by: Andrew Murray <andrew.murray@arm.com>
Reviewed-by: Suzuki K Poulose <suzuki.poulose@arm.com>
[Fixed capital letter for "use" in title]
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Link: https://lore.kernel.org/r/20190829202842.580-11-mathieu.poirier@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agocoresight: etm4x: Remove superfluous setting of os_unlock
Andrew Murray [Thu, 29 Aug 2019 20:28:34 +0000 (14:28 -0600)]
coresight: etm4x: Remove superfluous setting of os_unlock

In addition to unlocking the OS lock, etm4_os_unlock will also
set the os_unlock flag. Therefore let's avoid unnecessarily
setting os_unlock flag outside of this function.

Signed-off-by: Andrew Murray <andrew.murray@arm.com>
Reviewed-by: Suzuki K Poulose <suzuki.poulose@arm.com>
[Fixed capital letter for "remove" in the title]
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Link: https://lore.kernel.org/r/20190829202842.580-10-mathieu.poirier@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agocoresight: acpi: Static funnel support
Suzuki K Poulose [Thu, 29 Aug 2019 20:28:33 +0000 (14:28 -0600)]
coresight: acpi: Static funnel support

The ACPI bindings for CoreSight has been updated to add the device
id for non-programmable CoreSight funnels (aka static funnels) as of
v1.1 [0]. Add the ACPI id for static funnels in the driver.

[0] https://static.docs.arm.com/den0067/a/DEN0067_CoreSight_ACPI_1.1.pdf

Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Link: https://lore.kernel.org/r/20190829202842.580-9-mathieu.poirier@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agocoresight: Convert pr_warn to dev_warn for obsolete bindings
Suzuki K Poulose [Thu, 29 Aug 2019 20:28:32 +0000 (14:28 -0600)]
coresight: Convert pr_warn to dev_warn for obsolete bindings

We warn the users of obsolete bindings in the DT for coresight replicator
and funnel drivers. However we use pr_warn_once() which doesn't give a clue
about which device it is bound to. Let us use dev_warn_once() to give the
context.

Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Link: https://lore.kernel.org/r/20190829202842.580-8-mathieu.poirier@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agocoresight: tmc-etr: Check if non-secure access is enabled
Suzuki K Poulose [Thu, 29 Aug 2019 20:28:31 +0000 (14:28 -0600)]
coresight: tmc-etr: Check if non-secure access is enabled

CoreSight TMC-ETR must have the non-secure invasive debug access
enabled for use by self-hosted tracing. Without it, there is no
point in enabling the ETR. So, let us check it in the TMC_AUTHSTATUS
register and fail the probe if it is disabled.

Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Link: https://lore.kernel.org/r/20190829202842.580-7-mathieu.poirier@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agocoresight: tmc-etr: Handle memory errors
Suzuki K Poulose [Thu, 29 Aug 2019 20:28:30 +0000 (14:28 -0600)]
coresight: tmc-etr: Handle memory errors

We have so far ignored the memory errors, assuming that we have perfect
hardware and driver. Let us handle the memory errors reported by the
TMC ETR in status and truncate the buffer.

Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
[Removed ASCII smiley face from changelog]
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Link: https://lore.kernel.org/r/20190829202842.580-6-mathieu.poirier@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agocoresight: etr_buf: Consolidate refcount initialization
Suzuki K Poulose [Thu, 29 Aug 2019 20:28:29 +0000 (14:28 -0600)]
coresight: etr_buf: Consolidate refcount initialization

We now use refcounts for the etr_buf users. Let us initialize it
while we allocate it.

Cc: Mathieu Poirier <mathieu.poirier@linaro.org>
Signed-off-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Link: https://lore.kernel.org/r/20190829202842.580-5-mathieu.poirier@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agocoresight: cpu-debug: Add support for Qualcomm Kryo
Sai Prakash Ranjan [Thu, 29 Aug 2019 20:28:28 +0000 (14:28 -0600)]
coresight: cpu-debug: Add support for Qualcomm Kryo

Add support for coresight CPU debug module on Qualcomm
Kryo CPUs. This patch adds the UCI entries for Kryo CPUs
found on MSM8996 which shares the same PIDs as ETMs.

Without this, below error is observed on MSM8996:

[    5.429867] OF: graph: no port node found in /soc/debug@3810000
[    5.429938] coresight-etm4x: probe of 3810000.debug failed with error -22
[    5.435415] coresight-cpu-debug 3810000.debug: Coresight debug-CPU0 initialized
[    5.446474] OF: graph: no port node found in /soc/debug@3910000
[    5.448927] coresight-etm4x: probe of 3910000.debug failed with error -22
[    5.454681] coresight-cpu-debug 3910000.debug: Coresight debug-CPU1 initialized
[    5.487765] OF: graph: no port node found in /soc/debug@3a10000
[    5.488007] coresight-etm4x: probe of 3a10000.debug failed with error -22
[    5.493024] coresight-cpu-debug 3a10000.debug: Coresight debug-CPU2 initialized
[    5.501802] OF: graph: no port node found in /soc/debug@3b10000
[    5.512901] coresight-etm4x: probe of 3b10000.debug failed with error -22
[    5.513192] coresight-cpu-debug 3b10000.debug: Coresight debug-CPU3 initialized

Signed-off-by: Sai Prakash Ranjan <saiprakash.ranjan@codeaurora.org>
Tested-by: Leo Yan <leo.yan@linaro.org>
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Link: https://lore.kernel.org/r/20190829202842.580-4-mathieu.poirier@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agocoresight: etm4x: Add ETM PIDs for SDM845 and MSM8996
Sai Prakash Ranjan [Thu, 29 Aug 2019 20:28:27 +0000 (14:28 -0600)]
coresight: etm4x: Add ETM PIDs for SDM845 and MSM8996

Instead of overriding the peripheral id(PID) check in AMBA
by hardcoding them in DT, add the PIDs to the ETM4x driver.
Here we use Unique Component Identifier(UCI) for MSM8996
since the ETM and CPU debug module shares the same PIDs.
SDM845 does not support CPU debug module.

Signed-off-by: Sai Prakash Ranjan <saiprakash.ranjan@codeaurora.org>
Reviewed-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Link: https://lore.kernel.org/r/20190829202842.580-3-mathieu.poirier@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agocoresight: etm4x: Two function calls less
Markus Elfring [Thu, 29 Aug 2019 20:28:26 +0000 (14:28 -0600)]
coresight: etm4x: Two function calls less

Avoid an extra function call in two function implementations
by using a ternary operator instead of a conditional statement.

This issue was detected by using the Coccinelle software.

Signed-off-by: Markus Elfring <elfring@users.sourceforge.net>
Signed-off-by: Mathieu Poirier <mathieu.poirier@linaro.org>
Link: https://lore.kernel.org/r/20190829202842.580-2-mathieu.poirier@linaro.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agoMerge tag 'fpga-cvp-for-5.4' of git://git.kernel.org/pub/scm/linux/kernel/git/mdf...
Greg Kroah-Hartman [Tue, 3 Sep 2019 19:57:04 +0000 (21:57 +0200)]
Merge tag 'fpga-cvp-for-5.4' of git://git./linux/kernel/git/mdf/linux-fpga into char-misc-next

Moritz writes:

FPGA Manager changes for 5.4-rc1

Here is the second set of changes for the 5.4 merge window.

This patchset adds support for the v2 revision of Intel (Altera)'s CVP
parts including the Stratix 10.

All of this patches have been reviewed and been in the last few
linux-next releases without issues.

Signed-off-by: Moritz Fischer <mdf@kernel.org>
* tag 'fpga-cvp-for-5.4' of git://git.kernel.org/pub/scm/linux/kernel/git/mdf/linux-fpga:
  fpga: altera-cvp: Add Stratix10 (V2) Support
  fpga: altera-cvp: Preparation for V2 parts.
  fpga: altera-cvp: Discover Vendor Specific offset

5 years agoMerge tag 'thunderbolt-for-v5.4' of git://git.kernel.org/pub/scm/linux/kernel/git...
Greg Kroah-Hartman [Tue, 3 Sep 2019 19:56:00 +0000 (21:56 +0200)]
Merge tag 'thunderbolt-for-v5.4' of git://git./linux/kernel/git/westeri/thunderbolt into char-misc-next

Mika writes:

thunderbolt: Changes for v5.4 merge window

The biggest change is the addition of Intel Ice Lake integrated
Thunderbolt support. There are also a couple of smaller changes like
converting the driver to use better device property interface and use
correct format string in service key attribute.

* tag 'thunderbolt-for-v5.4' of git://git.kernel.org/pub/scm/linux/kernel/git/westeri/thunderbolt:
  ACPI / property: Add two new Thunderbolt property GUIDs to the list
  thunderbolt: Add support for Intel Ice Lake
  thunderbolt: Expose active parts of NVM even if upgrade is not supported
  thunderbolt: Hide switch attributes that are not set
  thunderbolt: Do not fail adding switch if some port is not implemented
  thunderbolt: Use 32-bit writes when writing ring producer/consumer
  thunderbolt: Move NVM upgrade support flag to struct icm
  thunderbolt: Correct path indices for PCIe tunnel
  thunderbolt: Show key using %*pE not %*pEp
  thunderbolt: Switch to use device_property_count_uXX()

5 years agoMerge tag 'phy-for-5.4' of git://git.kernel.org/pub/scm/linux/kernel/git/kishon/linux...
Greg Kroah-Hartman [Tue, 3 Sep 2019 19:49:11 +0000 (21:49 +0200)]
Merge tag 'phy-for-5.4' of git://git./linux/kernel/git/kishon/linux-phy into char-misc-next

Kishon writes:

phy: for 5.4

  *) Add a new PHY driver for Lantiq VRX200/ARX300 PCIe PHY
  *) Add missing of_node_put() to a bunch of drivers using
     for_each_available_child_of_node()
  *) Add RXAUI/PCIe/SATA/USB3 support in Marvell's Armada
     CP110 COMPHY
  *) Other misc fixes and cleanup

Signed-off-by: Kishon Vijay Abraham I <kishon@ti.com>
* tag 'phy-for-5.4' of git://git.kernel.org/pub/scm/linux/kernel/git/kishon/linux-phy: (30 commits)
  phy: marvell: phy-mvebu-cp110-comphy: rename instances of DLT
  phy: marvell: phy-mvebu-cp110-comphy: implement RXAUI support
  dt-bindings: pci: add PHY properties to Armada 7K/8K controller bindings
  dt-bindings: phy: Add Marvell COMPHY clocks
  phy: mvebu-cp110-comphy: Update comment about powering off all lanes at boot
  phy: mvebu-cp110-comphy: Add PCIe support
  phy: mvebu-cp110-comphy: Cosmetic change in a helper
  phy: mvebu-cp110-comphy: Add SATA support
  phy: mvebu-cp110-comphy: Add USB3 host/device support
  phy: mvebu-cp110-comphy: Allow non-Ethernet modes to be configured
  phy: mvebu-cp110-comphy: Rename the macro handling only Ethernet modes
  phy: mvebu-cp110-comphy: Add RXAUI support
  phy: mvebu-cp110-comphy: List already supported Ethernet modes
  phy: mvebu-cp110-comphy: Add SMC call support
  phy: mvebu-cp110-comphy: Explicitly initialize the lane submode
  phy: mvebu-cp110-comphy: Add clocks support
  phy-rockchip-inno-hdmi: Fix RK3328_TERM_RESISTOR_CALIB_SPEED_7_0's third value
  phy: qcom-qmp: Correct ready status, again
  phy: qualcomm: phy-qcom-qmp: Add of_node_put() before return
  phy: renesas: rcar-gen3-usb2: Disable clearing VBUS in over-current
  ...

5 years agoMerge tag 'icc-5.4-rc1' of https://git.linaro.org/people/georgi.djakov/linux into...
Greg Kroah-Hartman [Tue, 3 Sep 2019 19:47:37 +0000 (21:47 +0200)]
Merge tag 'icc-5.4-rc1' of https://git.linaro.org/people/georgi.djakov/linux into char-misc-next

Georgi writes:

interconnect patches for 5.4

Here are the interconnect driver updates for the 5.4-rc1 merge window.

- New feature is the path tagging support that helps with grouping and
aggregating the bandwidth requests into separate buckets based on a tag.
- The first user of the path tagging is the Qualcomm sdm845 driver that
now implements support for wake/sleep sets. This allows consumer drivers
to express their bandwidth needs for the different CPU power states.
- New interconnect driver for the qcs404 platforms and a driver that
communicates bandwidth requests with remote processor over shared memory.
- Cleanups and fixes.

Signed-off-by: Georgi Djakov <georgi.djakov@linaro.org>
* tag 'icc-5.4-rc1' of https://git.linaro.org/people/georgi.djakov/linux:
  drivers: qcom: Add BCM vote macro to header
  interconnect: qcom: remove COMPILE_TEST from CONFIG_INTERCONNECT_QCOM_QCS404
  interconnect: qcom: Add QCS404 interconnect provider driver
  interconnect: qcom: Add interconnect RPM over SMD driver
  dt-bindings: interconnect: Add Qualcomm QCS404 DT bindings
  interconnect: qcom: Add tagging and wake/sleep support for sdm845
  interconnect: Add pre_aggregate() callback
  interconnect: Add support for path tags

5 years agoMerge tag 'extcon-next-for-5.4' of git://git.kernel.org/pub/scm/linux/kernel/git...
Greg Kroah-Hartman [Tue, 3 Sep 2019 19:46:20 +0000 (21:46 +0200)]
Merge tag 'extcon-next-for-5.4' of git://git./linux/kernel/git/chanwoo/extcon into char-misc-next

Chanwoo writes:

Update extcon for 5.4

Detailed description for this pull request:
1. Clean up the and fix the minor issue of extcon provider driver
- extcon-arizona/max77843 replace the helper function
  with more correct helper function without operation changes.
- extcon-fsa9480 supports the FSA880 variant by adding the compatible name.
- extcon-arizona updates the dt-binding file for the readability.
- extcon-gpio initializes the interrupt flags according to active-low state.
- Clean up extcon-sm5502/axp288/adc-jack

* tag 'extcon-next-for-5.4' of git://git.kernel.org/pub/scm/linux/kernel/git/chanwoo/extcon:
  extcon: adc-jack: Remove dev_err() usage after platform_get_irq()
  extcon: axp288: Use for_each_set_bit() in axp288_extcon_log_rsi()
  extcon: axp288: Add missed error check
  extcon: sm5502: Add IRQ_ONESHOT
  extcon: gpio: Request reasonable interrupts
  extcon: arizona: Update binding example to use available defines
  extcon: fsa9480: Support the FSA880 variant
  extcon: extcon-max77843: convert to i2c_new_dummy_device
  extcon: arizona: Switch to use device_property_count_u32()

5 years agoMerge 5.3-rc7 into char-misc-next
Greg Kroah-Hartman [Mon, 2 Sep 2019 17:30:09 +0000 (19:30 +0200)]
Merge 5.3-rc7 into char-misc-next

We need the fixes in here as well for testing and merges

Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
5 years agoLinux 5.3-rc7
Linus Torvalds [Mon, 2 Sep 2019 16:57:40 +0000 (09:57 -0700)]
Linux 5.3-rc7

5 years agoMerge tag 'char-misc-5.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh...
Linus Torvalds [Mon, 2 Sep 2019 16:30:34 +0000 (09:30 -0700)]
Merge tag 'char-misc-5.3-rc7' of git://git./linux/kernel/git/gregkh/char-misc

Pull char/misc driver fixes from Greg KH:
 "Here are some small char and misc driver fixes for reported issues for
  5.3-rc7

  Also included in here is the documentation for how we are handling
  hardware issues under embargo that everyone has finally agreed on, as
  well as a MAINTAINERS update for the suckers who agreed to handle the
  LICENSES/ files.

  All of these have been in linux-next last week with no reported
  issues"

* tag 'char-misc-5.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/char-misc:
  fsi: scom: Don't abort operations for minor errors
  vmw_balloon: Fix offline page marking with compaction
  VMCI: Release resource if the work is already queued
  Documentation/process: Embargoed hardware security issues
  lkdtm/bugs: fix build error in lkdtm_EXHAUST_STACK
  mei: me: add Tiger Lake point LP device ID
  intel_th: pci: Add Tiger Lake support
  intel_th: pci: Add support for another Lewisburg PCH
  stm class: Fix a double free of stm_source_device
  MAINTAINERS: add entry for LICENSES and SPDX stuff
  fpga: altera-ps-spi: Fix getting of optional confd gpio

5 years agoMerge tag 'usb-5.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb
Linus Torvalds [Mon, 2 Sep 2019 16:15:30 +0000 (09:15 -0700)]
Merge tag 'usb-5.3-rc7' of git://git./linux/kernel/git/gregkh/usb

Pull USB fixes from Greg KH:
 "Here are some small USB fixes that have been in linux-next this past
  week for 5.3-rc7

  They fix the usual xhci, syzbot reports, and other small issues that
  have come up last week.

  All have been in linux-next with no reported issues"

* tag 'usb-5.3-rc7' of git://git.kernel.org/pub/scm/linux/kernel/git/gregkh/usb:
  USB: cdc-wdm: fix race between write and disconnect due to flag abuse
  usb: host: xhci: rcar: Fix typo in compatible string matching
  usb: host: xhci-tegra: Set DMA mask correctly
  USB: storage: ums-realtek: Whitelist auto-delink support
  USB: storage: ums-realtek: Update module parameter description for auto_delink_en
  usb: host: ohci: fix a race condition between shutdown and irq
  usb: hcd: use managed device resources
  typec: tcpm: fix a typo in the comparison of pdo_max_voltage
  usb-storage: Add new JMS567 revision to unusual_devs
  usb: chipidea: udc: don't do hardware access if gadget has stopped
  usbtmc: more sanity checking for packet size
  usb: udc: lpc32xx: silence fall-through warning

5 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Linus Torvalds [Mon, 2 Sep 2019 01:45:28 +0000 (18:45 -0700)]
Merge git://git./linux/kernel/git/netdev/net

Pull networking fixes from David Miller:

 1) Fix some length checks during OGM processing in batman-adv, from
    Sven Eckelmann.

 2) Fix regression that caused netfilter conntrack sysctls to not be
    per-netns any more. From Florian Westphal.

 3) Use after free in netpoll, from Feng Sun.

 4) Guard destruction of pfifo_fast per-cpu qdisc stats with
    qdisc_is_percpu_stats(), from Davide Caratti. Similar bug is fixed
    in pfifo_fast_enqueue().

 5) Fix memory leak in mld_del_delrec(), from Eric Dumazet.

 6) Handle neigh events on internal ports correctly in nfp, from John
    Hurley.

 7) Clear SKB timestamp in NF flow table code so that it does not
    confuse fq scheduler. From Florian Westphal.

 8) taprio destroy can crash if it is invoked in a failure path of
    taprio_init(), because the list head isn't setup properly yet and
    the list del is unconditional. Perform the list add earlier to
    address this. From Vladimir Oltean.

 9) Make sure to reapply vlan filters on device up, in aquantia driver.
    From Dmitry Bogdanov.

10) sgiseeq driver releases DMA memory using free_page() instead of
    dma_free_attrs(). From Christophe JAILLET.

* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (58 commits)
  net: seeq: Fix the function used to release some memory in an error handling path
  enetc: Add missing call to 'pci_free_irq_vectors()' in probe and remove functions
  net: bcmgenet: use ethtool_op_get_ts_info()
  tc-testing: don't hardcode 'ip' in nsPlugin.py
  net: dsa: microchip: add KSZ8563 compatibility string
  dt-bindings: net: dsa: document additional Microchip KSZ8563 switch
  net: aquantia: fix out of memory condition on rx side
  net: aquantia: linkstate irq should be oneshot
  net: aquantia: reapply vlan filters on up
  net: aquantia: fix limit of vlan filters
  net: aquantia: fix removal of vlan 0
  net/sched: cbs: Set default link speed to 10 Mbps in cbs_set_port_rate
  taprio: Set default link speed to 10 Mbps in taprio_set_picos_per_byte
  taprio: Fix kernel panic in taprio_destroy
  net: dsa: microchip: fill regmap_config name
  rxrpc: Fix lack of conn cleanup when local endpoint is cleaned up [ver #2]
  net: stmmac: dwmac-rk: Don't fail if phy regulator is absent
  amd-xgbe: Fix error path in xgbe_mod_init()
  netfilter: nft_meta_bridge: Fix get NFT_META_BRI_IIFVPROTO in network byteorder
  mac80211: Correctly set noencrypt for PAE frames
  ...

5 years agonet: seeq: Fix the function used to release some memory in an error handling path
Christophe JAILLET [Sat, 31 Aug 2019 07:17:51 +0000 (09:17 +0200)]
net: seeq: Fix the function used to release some memory in an error handling path

In commit 99cd149efe82 ("sgiseeq: replace use of dma_cache_wback_inv"),
a call to 'get_zeroed_page()' has been turned into a call to
'dma_alloc_coherent()'. Only the remove function has been updated to turn
the corresponding 'free_page()' into 'dma_free_attrs()'.
The error hndling path of the probe function has not been updated.

Fix it now.

Rename the corresponding label to something more in line.

Fixes: 99cd149efe82 ("sgiseeq: replace use of dma_cache_wback_inv")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Reviewed-by: Thomas Bogendoerfer <tbogendoerfer@suse.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
5 years agoMerge branch 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 1 Sep 2019 18:21:57 +0000 (11:21 -0700)]
Merge branch 'x86-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull x86 fixes from Thomas Gleixner:
 "A set of fixes for x86:

   - Fix the bogus detection of 32bit user mode for uretprobes which
     caused corruption of the user return address resulting in
     application crashes. In the uprobes handler in_ia32_syscall() is
     obviously always returning false on a 64bit kernel. Use
     user_64bit_mode() instead which works correctly.

   - Prevent large page splitting when ftrace flips RW/RO on the kernel
     text which caused iTLB performance issues. Ftrace wants to be
     converted to text_poke() which avoids the problem, but for now
     allow large page preservation in the static protections check when
     the change request spawns a full large page.

   - Prevent arch_dynirq_lower_bound() from returning 0 when the IOAPIC
     is configured via device tree. In the device tree case the GSI 1:1
     mapping is meaningless therefore the lower bound which protects the
     GSI range on ACPI machines is irrelevant. Return the lower bound
     which the core hands to the function instead of blindly returning 0
     which causes the core to allocate the invalid virtual interupt
     number 0 which in turn prevents all drivers from allocating and
     requesting an interrupt.

   - Remove the bogus initialization of LDR and DFR in the 32bit bigsmp
     APIC driver. That uses physical destination mode where LDR/DFR are
     ignored, but the initialization and the missing clear of LDR caused
     the APIC to be left in a inconsistent state on kexec/reboot.

   - Clear LDR when clearing the APIC registers so the APIC is in a well
     defined state.

   - Initialize variables proper in the find_trampoline_placement()
     code.

   - Silence GCC( build warning for the real mode part of the build"

* 'x86-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  x86/mm/cpa: Prevent large page split when ftrace flips RW on kernel text
  x86/build: Add -Wnoaddress-of-packed-member to REALMODE_CFLAGS, to silence GCC9 build warning
  x86/boot/compressed/64: Fix missing initialization in find_trampoline_placement()
  x86/apic: Include the LDR when clearing out APIC registers
  x86/apic: Do not initialize LDR and DFR for bigsmp
  uprobes/x86: Fix detection of 32-bit user mode
  x86/apic: Fix arch_dynirq_lower_bound() bug for DT enabled machines

5 years agoMerge branch 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel...
Linus Torvalds [Sun, 1 Sep 2019 18:09:42 +0000 (11:09 -0700)]
Merge branch 'perf-urgent-for-linus' of git://git./linux/kernel/git/tip/tip

Pull perf fixes from Thomas Gleixner:
 "Two fixes for perf x86 hardware implementations:

   - Restrict the period on Nehalem machines to prevent perf from
     hogging the CPU

   - Prevent the AMD IBS driver from overwriting the hardwre controlled
     and pre-seeded reserved bits (0-6) in the count register which
     caused a sample bias for dispatched micro-ops"

* 'perf-urgent-for-linus' of git://git.kernel.org/pub/scm/linux/kernel/git/tip/tip:
  perf/x86/amd/ibs: Fix sample bias for dispatched micro-ops
  perf/x86/intel: Restrict period on Nehalem

5 years agoMerge branch 'turbostat' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux
Linus Torvalds [Sun, 1 Sep 2019 17:39:25 +0000 (10:39 -0700)]
Merge branch 'turbostat' of git://git./linux/kernel/git/lenb/linux

Pull turbostat updates from Len Brown:
 "User-space turbostat (and x86_energy_perf_policy) patches.

  They are primarily bug fixes from users"

* 'turbostat' of git://git.kernel.org/pub/scm/linux/kernel/git/lenb/linux:
  tools/power turbostat: update version number
  tools/power turbostat: Add support for Hygon Fam 18h (Dhyana) RAPL
  tools/power turbostat: Fix caller parameter of get_tdp_amd()
  tools/power turbostat: Fix CPU%C1 display value
  tools/power turbostat: do not enforce 1ms
  tools/power turbostat: read from pipes too
  tools/power turbostat: Add Ice Lake NNPI support
  tools/power turbostat: rename has_hsw_msrs()
  tools/power turbostat: Fix Haswell Core systems
  tools/power turbostat: add Jacobsville support
  tools/power turbostat: fix buffer overrun
  tools/power turbostat: fix file descriptor leaks
  tools/power turbostat: fix leak of file descriptor on error return path
  tools/power turbostat: Make interval calculation per thread to reduce jitter
  tools/power turbostat: remove duplicate pc10 column
  tools/power x86_energy_perf_policy: Fix argument parsing
  tools/power: Fix typo in man page
  tools/power/x86: Enable compiler optimisations and Fortify by default
  tools/power x86_energy_perf_policy: Fix "uninitialized variable" warnings at -O2

5 years agoenetc: Add missing call to 'pci_free_irq_vectors()' in probe and remove functions
Christophe JAILLET [Fri, 30 Aug 2019 20:23:12 +0000 (22:23 +0200)]
enetc: Add missing call to 'pci_free_irq_vectors()' in probe and remove functions

Call to 'pci_free_irq_vectors()' are missing both in the error handling
path of the probe function, and in the remove function.
Add them.

Fixes: 19971f5ea0ab ("enetc: add PTP clock driver")
Signed-off-by: Christophe JAILLET <christophe.jaillet@wanadoo.fr>
Signed-off-by: David S. Miller <davem@davemloft.net>
5 years agonet: bcmgenet: use ethtool_op_get_ts_info()
Ryan M. Collins [Fri, 30 Aug 2019 18:49:55 +0000 (14:49 -0400)]
net: bcmgenet: use ethtool_op_get_ts_info()

This change enables the use of SW timestamping on the Raspberry Pi 4.

bcmgenet's transmit function bcmgenet_xmit() implements software
timestamping. However the SOF_TIMESTAMPING_TX_SOFTWARE capability was
missing and only SOF_TIMESTAMPING_RX_SOFTWARE was announced. By using
ethtool_ops bcmgenet_ethtool_ops() as get_ts_info(), the
SOF_TIMESTAMPING_TX_SOFTWARE capability is announced.

Similar to commit a8f5cb9e7991 ("smsc95xx: use ethtool_op_get_ts_info()")

Signed-off-by: Ryan M. Collins <rmc032@bucknell.edu>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Acked-by: Doug Berger <opendmb@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
5 years agotc-testing: don't hardcode 'ip' in nsPlugin.py
Davide Caratti [Fri, 30 Aug 2019 16:51:47 +0000 (18:51 +0200)]
tc-testing: don't hardcode 'ip' in nsPlugin.py

the following tdc test fails on Fedora:

 # ./tdc.py -e 2638
  -- ns/SubPlugin.__init__
 Test 2638: Add matchall and try to get it
 -----> prepare stage *** Could not execute: "$TC qdisc add dev $DEV1 clsact"
 -----> prepare stage *** Error message: "/bin/sh: ip: command not found"
 returncode 127; expected [0]
 -----> prepare stage *** Aborting test run.

Let nsPlugin.py use the 'IP' variable introduced with commit 92c1a19e2fb9
("tc-tests: added path to ip command in tdc"), so that the path to 'ip' is
correctly resolved to the value we have in tdc_config.py.

 # ./tdc.py -e 2638
  -- ns/SubPlugin.__init__
 Test 2638: Add matchall and try to get it
 All test results:
 1..1
 ok 1 2638 - Add matchall and try to get it

Fixes: 489ce2f42514 ("tc-testing: Restore original behaviour for namespaces in tdc")
Reported-by: Hangbin Liu <liuhangbin@gmail.com>
Signed-off-by: Davide Caratti <dcaratti@redhat.com>
Acked-by: Nicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
5 years agoMerge branch 'net-dsa-microchip-add-KSZ8563-support'
David S. Miller [Sun, 1 Sep 2019 06:36:37 +0000 (23:36 -0700)]
Merge branch 'net-dsa-microchip-add-KSZ8563-support'

Razvan Stefanescu says:

====================
net: dsa: microchip: add KSZ8563 support

This patchset adds compatibility string for the KSZ8563 switch.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
5 years agonet: dsa: microchip: add KSZ8563 compatibility string
Razvan Stefanescu [Fri, 30 Aug 2019 07:52:02 +0000 (10:52 +0300)]
net: dsa: microchip: add KSZ8563 compatibility string

It is a 3-Port 10/100 Ethernet Switch with 1588v2 PTP.

Signed-off-by: Razvan Stefanescu <razvan.stefanescu@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
5 years agodt-bindings: net: dsa: document additional Microchip KSZ8563 switch
Razvan Stefanescu [Fri, 30 Aug 2019 07:52:01 +0000 (10:52 +0300)]
dt-bindings: net: dsa: document additional Microchip KSZ8563 switch

It is a 3-Port 10/100 Ethernet Switch with 1588v2 PTP.

Signed-off-by: Razvan Stefanescu <razvan.stefanescu@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
5 years agoMerge branch 'net-aquantia-fixes-on-vlan-filters-and-other-conditions'
David S. Miller [Sun, 1 Sep 2019 02:07:17 +0000 (19:07 -0700)]
Merge branch 'net-aquantia-fixes-on-vlan-filters-and-other-conditions'

Igor Russkikh says:

====================
net: aquantia: fixes on vlan filters and other conditions

Here is a set of various bug fixes related to vlan filter offload and
two other rare cases.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
5 years agonet: aquantia: fix out of memory condition on rx side
Dmitry Bogdanov [Fri, 30 Aug 2019 12:08:38 +0000 (12:08 +0000)]
net: aquantia: fix out of memory condition on rx side

On embedded environments with hard memory limits it is a normal although
rare case when skb can't be allocated on rx part under high traffic.

In such OOM cases napi_complete_done() was not called.
So the napi object became in an invalid state like it is "scheduled".
Kernel do not re-schedules the poll of that napi object.

Consequently, kernel can not remove that object the system hangs on
`ifconfig down` waiting for a poll.

We are fixing this by gracefully closing napi poll routine with correct
invocation of napi_complete_done.

This was reproduced with artificially failing the allocation of skb to
simulate an "out of memory" error case and check that traffic does
not get stuck.

Fixes: 970a2e9864b0 ("net: ethernet: aquantia: Vector operations")
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
Signed-off-by: Dmitry Bogdanov <dmitry.bogdanov@aquantia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
5 years agonet: aquantia: linkstate irq should be oneshot
Igor Russkikh [Fri, 30 Aug 2019 12:08:36 +0000 (12:08 +0000)]
net: aquantia: linkstate irq should be oneshot

Declaring threaded irq handler should also indicate the irq is
oneshot. It is oneshot indeed, because HW implements irq automasking
on trigger.

Not declaring this causes some kernel configurations to fail
on interface up, because request_threaded_irq returned an err code.

The issue was originally hidden on normal x86_64 configuration with
latest kernel, because depending on interrupt controller, irq driver
added ONESHOT flag on its own.

Issue was observed on older kernels (4.14) where no such logic exists.

Fixes: 4c83f170b3ac ("net: aquantia: link status irq handling")
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
Reported-by: Michael Symolkin <Michael.Symolkin@aquantia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
5 years agonet: aquantia: reapply vlan filters on up
Dmitry Bogdanov [Fri, 30 Aug 2019 12:08:35 +0000 (12:08 +0000)]
net: aquantia: reapply vlan filters on up

In case of device reconfiguration the driver may reset the device invisible
for other modules, vlan module in particular. So vlans will not be
removed&created and vlan filters will not be configured in the device.
The patch reapplies the vlan filters at device start.

Fixes: 7975d2aff5afb ("net: aquantia: add support of rx-vlan-filter offload")
Signed-off-by: Dmitry Bogdanov <dmitry.bogdanov@aquantia.com>
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
5 years agonet: aquantia: fix limit of vlan filters
Dmitry Bogdanov [Fri, 30 Aug 2019 12:08:33 +0000 (12:08 +0000)]
net: aquantia: fix limit of vlan filters

Fix a limit condition of vlans on the interface before setting vlan
promiscuous mode

Fixes: 48dd73d08d4dd ("net: aquantia: fix vlans not working over bridged network")
Signed-off-by: Dmitry Bogdanov <dmitry.bogdanov@aquantia.com>
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
5 years agonet: aquantia: fix removal of vlan 0
Dmitry Bogdanov [Fri, 30 Aug 2019 12:08:30 +0000 (12:08 +0000)]
net: aquantia: fix removal of vlan 0

Due to absence of checking against the rx flow rule when vlan 0 is being
removed, the other rule could be removed instead of the rule with vlan 0

Fixes: 7975d2aff5afb ("net: aquantia: add support of rx-vlan-filter offload")
Signed-off-by: Dmitry Bogdanov <dmitry.bogdanov@aquantia.com>
Signed-off-by: Igor Russkikh <igor.russkikh@aquantia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>