On Thu, Jul 10, 2014 at 03:38:33PM -0700, Joe Perches wrote:
> On Fri, 2014-07-11 at 00:50 +0300, Oded Gabbay wrote:
> > This patch adds the interface between the radeon driver and the kfd
> > driver. The interface implementation is contained in
> > radeon_kfd.c and radeon_kfd.h.
> []
> > include/
On Fri, Jul 11, 2014 at 12:50:05AM +0300, Oded Gabbay wrote:
> This patch adds a new interface to kfd2kgd_calls structure so that
> the kfd driver could get the virtual ram size of a specific
> radeon device.
>
> Signed-off-by: Oded Gabbay
What is vmem_size ? This need to be documented. I assume
https://bugzilla.kernel.org/show_bug.cgi?id=79071
Jonathan Howard changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|---
https://bugzilla.kernel.org/show_bug.cgi?id=79051
Jonathan Howard changed:
What|Removed |Added
Status|NEW |RESOLVED
Resolution|---
On Fri, Jul 11, 2014 at 12:50:06AM +0300, Oded Gabbay wrote:
> This patch adds new interfaces to kfd2kgd_calls structure.
>
> The new interfaces allow the kfd driver to :
>
> 1. Allocated video memory through the radeon driver
> 2. Map and unmap video memory with GPUVM through the radeon driver
>
On Fri, Jul 11, 2014 at 12:50:07AM +0300, Oded Gabbay wrote:
> This patch adds a new interface to kfd2kgd_calls structure, which
> allows the kfd to lock and unlock the srbm_gfx_cntl register
Why does kfd needs to lock this register if kfd can not access
any of those register ? This sounds broken
On Fri, Jul 11, 2014 at 12:50:08AM +0300, Oded Gabbay wrote:
> The KFD driver should be loaded when the radeon driver is loaded and
> should be finalized when the radeon driver is removed.
>
> This patch adds a function call to initialize kfd from radeon_init
> and a function call to finalize kfd
On Fri, Jul 11, 2014 at 12:50:09AM +0300, Oded Gabbay wrote:
> This patch adds the code base of the hsa driver for
> AMD's GPUs.
>
> This driver is called kfd.
>
> This initial version supports the first HSA chip, Kaveri.
>
> This driver is located in a new directory structure under drivers/gpu.
>-Original Message-
>From: dri-devel [mailto:dri-devel-bounces at lists.freedesktop.org] On Behalf
>Of Alex Deucher
>Sent: Friday, July 11, 2014 12:23 PM
>To: Koenig, Christian
>Cc: Oded Gabbay; Lewycky, Andrew; LKML; Maling list - DRI developers;
>Deucher, Alexander
>Subject: Re: [PATCH
On Fri, 2014-07-11 at 13:04 -0400, Jerome Glisse wrote:
> On Fri, Jul 11, 2014 at 12:50:09AM +0300, Oded Gabbay wrote:
[]
> > +static long kfd_ioctl(struct file *, unsigned int, unsigned long);
>
> Nitpick, avoid unsigned int just use unsigned.
I suggest unsigned int is much more common (and bett
On Fri, Jul 11, 2014 at 7:04 PM, Jerome Glisse wrote:
> Are we to assume that for eternity this will not work on iommu that do support
> PASID/ATS but are not from AMD ? If it was an APU specific function i would
> understand but it seems that the IOMMU API needs to grow. I am pretty sure
> Intel
Hello Fabio,
On 11 Jul 12:08 PM, Fabio Estevam wrote:
> On Fri, Jul 11, 2014 at 11:18 AM, Ezequiel Garcia
> wrote:
> > In order to support the "enable GPIO" available in many panel devices,
> > this commit adds a proper devicetree binding.
> >
> > By providing an enable GPIO in the devicetree, th
Checking... we shouldn't need to call the lock from kfd any more.We should be
able to do any required locking in radeon kgd code.
>-Original Message-
>From: Jerome Glisse [mailto:j.glisse at gmail.com]
>Sent: Friday, July 11, 2014 12:35 PM
>To: Oded Gabbay
>Cc: David Airlie; Deucher, Alex
n HTML attachment was scrubbed...
URL:
<http://lists.freedesktop.org/archives/dri-devel/attachments/20140711/bd9bbc2b/attachment.html>
>-Original Message-
>From: Jerome Glisse [mailto:j.glisse at gmail.com]
>Sent: Friday, July 11, 2014 1:04 PM
>To: Oded Gabbay
>Cc: David Airlie; Deucher, Alexander; linux-kernel at vger.kernel.org; dri-
>devel at lists.freedesktop.org; Bridgman, John; Lewycky, Andrew; Joerg
>Roedel; Gabba
On Fri, Jul 11, 2014 at 06:02:39PM +, Bridgman, John wrote:
> >From: Jerome Glisse [mailto:j.glisse at gmail.com]
> >Sent: Friday, July 11, 2014 1:04 PM
> >To: Oded Gabbay
> >Cc: David Airlie; Deucher, Alexander; linux-kernel at vger.kernel.org; dri-
> >devel at lists.freedesktop.org; Bridgman,
ttp://lists.freedesktop.org/archives/dri-devel/attachments/20140711/98ead57b/attachment.html>
On Fri, Jul 11, 2014 at 12:50:11AM +0300, Oded Gabbay wrote:
> This patch adds the code base of the scheduler, which handles queue
> creation, deletion and scheduling on the CP of the GPU.
>
> Signed-off-by: Oded Gabbay
I would rather see all this squashed, this gave feeling that driver
can acce
>-Original Message-
>From: Jerome Glisse [mailto:j.glisse at gmail.com]
>Sent: Friday, July 11, 2014 2:11 PM
>To: Bridgman, John
>Cc: Oded Gabbay; David Airlie; Deucher, Alexander; linux-
>kernel at vger.kernel.org; dri-devel at lists.freedesktop.org; Lewycky, Andrew;
>Joerg Roedel; Gabba
On Fri, Jul 11, 2014 at 12:50:12AM +0300, Oded Gabbay wrote:
> This patch adds the kfd mmap handler that maps the physical address
> of a doorbell page to a user-space virtual address. That virtual address
> belongs to the process that uses the doorbell page.
>
> This mmap handler is called only f
On Fri, Jul 11, 2014 at 06:46:30PM +, Bridgman, John wrote:
> >From: Jerome Glisse [mailto:j.glisse at gmail.com]
> >Sent: Friday, July 11, 2014 2:11 PM
> >To: Bridgman, John
> >Cc: Oded Gabbay; David Airlie; Deucher, Alexander; linux-
> >kernel at vger.kernel.org; dri-devel at lists.freedeskto
>-Original Message-
>From: Jerome Glisse [mailto:j.glisse at gmail.com]
>Sent: Friday, July 11, 2014 2:52 PM
>To: Bridgman, John
>Cc: Oded Gabbay; David Airlie; Deucher, Alexander; linux-
>kernel at vger.kernel.org; dri-devel at lists.freedesktop.org; Lewycky, Andrew;
>Joerg Roedel; Gabba
On Fri, Jul 11, 2014 at 12:50:13AM +0300, Oded Gabbay wrote:
> This patch adds 2 new IOCTL to kfd driver.
>
> The first IOCTL is KFD_IOC_CREATE_QUEUE that is used by the user-mode
> application to create a compute queue on the GPU.
>
> The second IOCTL is KFD_IOC_DESTROY_QUEUE that is used by the
On Fri, Jul 11, 2014 at 06:56:12PM +, Bridgman, John wrote:
> >From: Jerome Glisse [mailto:j.glisse at gmail.com]
> >Sent: Friday, July 11, 2014 2:52 PM
> >To: Bridgman, John
> >Cc: Oded Gabbay; David Airlie; Deucher, Alexander; linux-
> >kernel at vger.kernel.org; dri-devel at lists.freedeskto
On Fri, 2014-07-11 at 15:22 -0400, Jerome Glisse wrote:
> Just to be explicit, my point is that is you claim GPL in MODULE_LICENSE
> then this is a GPL licensed code, if you claim GPL with additional rights
> than this is dual licensed code. This is how i read and interpret this
> with additional r
On Fri, Jul 11, 2014 at 08:29:56AM +0200, Daniel Vetter wrote:
> On Thu, Jul 10, 2014 at 09:08:24PM +, Tian, Kevin wrote:
> > actually I'm curious whether it's still necessary to __detect__ PCH. Could
> > we assume a 1:1 mapping between GPU and PCH, e.g. BDW already hard
> > code the knowledge:
On Fri, Jul 11, 2014 at 12:50:15AM +0300, Oded Gabbay wrote:
> This patch adds the interrupt handling module, in kfd_interrupt.c,
> and its related members in different data structures to the KFD
> driver.
>
> The KFD interrupt module maintains an internal interrupt ring per kfd
> device. The inte
--- next part --
An HTML attachment was scrubbed...
URL:
<http://lists.freedesktop.org/archives/dri-devel/attachments/20140711/ec0fef18/attachment.html>
On Fri, Jul 11, 2014 at 12:53:48AM +0300, Oded Gabbay wrote:
> From: Evgeny Pinchuk
>
> Implemented new IOCTL to query the CPU and GPU clock counters.
>
> Signed-off-by: Evgeny Pinchuk
> Signed-off-by: Oded Gabbay
> ---
> drivers/gpu/hsa/radeon/kfd_chardev.c | 37
> ++
On Fri, Jul 11, 2014 at 12:54:00AM +0300, Oded Gabbay wrote:
> From: Alexey Skidanov
>
> Added apertures initialization and appropriate ioctl
What is process aperture and what it is use for ? This is a very
cryptic commit message.
Cheers,
J?r?me
>
> Signed-off-by: Alexey Skidanov
> Signed-of
On Fri, Jul 11, 2014 at 12:50:13AM +0300, Oded Gabbay wrote:
> This patch adds 2 new IOCTL to kfd driver.
>
> The first IOCTL is KFD_IOC_CREATE_QUEUE that is used by the user-mode
> application to create a compute queue on the GPU.
>
> The second IOCTL is KFD_IOC_DESTROY_QUEUE that is used by the
On Thu, Jul 10, 2014 at 10:51:29PM +, Gabbay, Oded wrote:
> On Thu, 2014-07-10 at 18:24 -0400, Jerome Glisse wrote:
> > On Fri, Jul 11, 2014 at 12:45:27AM +0300, Oded Gabbay wrote:
> > > This patch set implements a Heterogeneous System Architecture
> > > (HSA) driver
> > > for radeon-family
David,
Please incorporate the latest Armada DRM updates, which can be found at:
git://ftp.arm.linux.org.uk/~rmk/linux-arm.git drm-armada-devel
with SHA1 9611cb93fa65dde199f4f888bd034ffc80c7adf0, based on v3.16-rc3.
This pull includes the component helpers which have been merged into
Greg's dr
> From: Konrad Rzeszutek Wilk [mailto:konrad.wilk at oracle.com]
> Sent: Friday, July 11, 2014 12:42 PM
>
> On Fri, Jul 11, 2014 at 08:29:56AM +0200, Daniel Vetter wrote:
> > On Thu, Jul 10, 2014 at 09:08:24PM +, Tian, Kevin wrote:
> > > actually I'm curious whether it's still necessary to __d
On Thu, Jul 10, 2014 at 11:55:26AM +0200, Thierry Reding wrote:
> On Wed, Mar 26, 2014 at 09:32:47AM +0100, Thierry Reding wrote:
> > On Tue, Mar 25, 2014 at 07:01:10PM +0100, Sam Ravnborg wrote:
> > > >
> > > > There are two things that don't work too well with this. First this
> > > > causes the
On Tue, Jul 08, 2014 at 10:31:59AM +0530, sonika.jindal at intel.com wrote:
> From: Ville Syrj?l?
>
> Sprite planes support 180 degree rotation. The lower layers are now in
> place, so hook in the standard rotation property to expose the feature
> to the users.
>
> v2: Moving rotation_property t
is enabled.
With the good old profile method the system is stable.
--
You are receiving this mail because:
You are the assignee for the bug.
-- next part --
An HTML attachment was scrubbed...
URL:
<http://lists.freedesktop.org/archives/dri-devel/attachments/2014
does indeed run or keep it open for performance
issues?
--
You are receiving this mail because:
You are the assignee for the bug.
-- next part --
An HTML attachment was scrubbed...
URL:
<http://lists.freedesktop.org/archives/dri-devel/attachments/20140711/07c7710d/attachment.html>
>From: Ilyes Gouta [mailto:ilyes.gouta at gmail.com]
>Sent: Friday, July 11, 2014 2:00 PM
>To: Bridgman, John
>Cc: Alex Deucher; Koenig, Christian; Oded Gabbay; Deucher, Alexander; Lewycky,
>Andrew; LKML; Maling list - DRI developers
>Subject: Re: [PATCH 02/83] drm/radeon: reduce number of free V
On 2014-07-10 11:33, Dan Carpenter wrote:
> On Wed, Jul 09, 2014 at 11:31:45PM +0200, Thomas Hellstr?m wrote:
>>> Speaking of verbose, all the likely/unlikely annotations should be
>>> removed.
>> Is this your personal opinion or has there been some kind of kernel
>> developer agreement not to add
On Fri, Jul 11, 2014 at 12:53:26AM +0300, Oded Gabbay wrote:
> mm/rmap.c | 8 ++--
> 1 file changed, 6 insertions(+), 2 deletions(-)
>
> diff --git a/mm/rmap.c b/mm/rmap.c
> index 196cd0c..73d4c3d 100644
> --- a/mm/rmap.c
> +++ b/mm/rmap.c
> @@ -1231,13 +1231,17 @@ static int try_to_unmap_one
This patch adds new interfaces to kfd2kgd_calls structure.
The new interfaces allow the kfd driver to :
1. Allocated video memory through the radeon driver
2. Map and unmap video memory with GPUVM through the radeon driver
3. Map and unmap system memory with GPUVM through the radeon driver
Signe
The KFD driver should be loaded when the radeon driver is loaded and
should be finalized when the radeon driver is removed.
This patch adds a function call to initialize kfd from radeon_init
and a function call to finalize kfd from radeon_exit.
If the KFD driver is not present in the system, the
This patch adds the code base of the scheduler, which handles queue
creation, deletion and scheduling on the CP of the GPU.
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/Makefile | 3 +-
drivers/gpu/hsa/radeon/cik_regs.h | 213 +++
drivers/gpu/hsa/radeon/k
When radeon handles interrupts for cik, propogate this interrupt to kfd.
Signed-off-by: Oded Gabbay
---
drivers/gpu/drm/radeon/cik.c| 4
drivers/gpu/drm/radeon/radeon_kfd.c | 6 ++
2 files changed, 10 insertions(+)
diff --git a/drivers/gpu/drm/radeon/cik.c b/drivers/gpu/drm/rad
From: Evgeny Pinchuk
The assignment of OEMID from the CRAT table is into a 64 variable. The OEMID is
48bit wide in the CRAT.
This fix makes sure that only 48bit are assigned for the OEMID value from the
CRAT table.
Signed-off-by: Evgeny Pinchuk
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa
From: Evgeny Pinchuk
Fixed wrong reporting of timestamps in kfd_ioctl_get_clock_counters.
Signed-off-by: Evgeny Pinchuk
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/kfd_chardev.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/hsa/radeon/kfd_charde
Radeon and KFD share the doorbell aperture.
Radeon sets it up, takes the doorbells required for its own rings
and reports the setup to KFD.
Radeon reserved doorbells are at the start of the doorbell aperture.
Signed-off-by: Oded Gabbay
---
drivers/gpu/drm/radeon/radeon.h| 4
driver
This patch adds the interface between the radeon driver and the kfd
driver. The interface implementation is contained in
radeon_kfd.c and radeon_kfd.h.
The interface itself is represented by a pointer to struct
kfd_dev. The pointer is located inside radeon_device structure.
Signed-off-by: Oded Ga
This patch adds a new interface to kfd2kgd_calls structure, which
allows the kfd to lock and unlock the srbm_gfx_cntl register
Signed-off-by: Oded Gabbay
---
drivers/gpu/drm/radeon/radeon_kfd.c | 20
include/linux/radeon_kfd.h | 4
2 files changed, 24 insertio
This patch adds the code base of the hsa driver for
AMD's GPUs.
This driver is called kfd.
This initial version supports the first HSA chip, Kaveri.
This driver is located in a new directory structure under drivers/gpu.
Signed-off-by: Oded Gabbay
---
drivers/Kconfig|
This patch adds initialization of the doorbell aperture when
initializing a kfd device.
It also adds a call to unmap the doorbell when a process unbinds
from the kfd
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/Makefile | 3 +-
drivers/gpu/hsa/radeon/kfd_device.c | 2 +
drive
This patch adds the isr function to the KFD scheduler code. This
function us called from the kgd2kfd_interrupt function which is
an interrupt-context function.
The purpose of the isr function is to determine whether the interrupt
that arrived is interesting, i.e. some action need to be taken.
Sig
This patch modifies the scheduler code to use interrupts to handle the
deactivation of queues. We prefer to use interrupts because the
deactivation could take a long time since we need to wait for the
wavefront to finish executing before deactivating the queue.
There is an array of waitqueues, eac
This patch add calls to initialize and finalize the KFD interrupt
module.
The calls are done per device initialize/finalize inside the kgd-->kfd
interface.
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/cik_regs.h | 1 +
drivers/gpu/hsa/radeon/kfd_device.c | 10 --
2 files cha
HSA radeon driver (kfd) should set interrupts for pipes 1-7.
Signed-off-by: Oded Gabbay
---
drivers/gpu/drm/radeon/cik.c | 71 +---
1 file changed, 1 insertion(+), 70 deletions(-)
diff --git a/drivers/gpu/drm/radeon/cik.c b/drivers/gpu/drm/radeon/cik.c
in
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/kfd_sched_cik_static.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/gpu/hsa/radeon/kfd_sched_cik_static.c
b/drivers/gpu/hsa/radeon/kfd_sched_cik_static.c
index 3c3e7d6..5bfde5c 100644
--- a/drivers/gpu/hsa/radeo
From: Evgeny Pinchuk
Implemented new IOCTL to query the CPU and GPU clock counters.
Signed-off-by: Evgeny Pinchuk
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/kfd_chardev.c | 37
include/uapi/linux/kfd_ioctl.h | 9 +
2 files changed
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/kfd_chardev.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)
diff --git a/drivers/gpu/hsa/radeon/kfd_chardev.c
b/drivers/gpu/hsa/radeon/kfd_chardev.c
index d6fa980..dba6084 100644
--- a/drivers/gpu/hsa/radeon/kfd_chardev.c
+++ b
From: Evgeny Pinchuk
Adding new function to the interface used by kfd.
The new function retrieves the max engine clock speed.
Signed-off-by: Evgeny Pinchuk
Signed-off-by: Oded Gabbay
---
drivers/gpu/drm/radeon/radeon_kfd.c | 11 +++
include/linux/radeon_kfd.h | 2 ++
2 files
From: Evgeny Pinchuk
Adding support for CPU and GPU max clock speeds in node properties.
Signed-off-by: Evgeny Pinchuk
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/kfd_topology.c | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/hsa/radeon/kfd_topo
From: Ben Goz
The queue module enables allocating and initializing queues uniformly.
The hw_pointer_store module handles allocation and assignment of read and write
pointers to user HSA queues.
Signed-off-by: Ben Goz
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/Makefile
From: Ben Goz
The kernel queue module enables the KFD to establish kernel queues, not exposed
to user space. The kernel queues are used for HIQ (HSA Interface Queue) and DIQ
(Debug Interface Queue) operations.
Signed-off-by: Ben Goz
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/Makefi
From: Evgeny Pinchuk
Introducing IOCTL implementation for controlling exclusive access to performace
counters.
The exclusive access is per GPU device.
Signed-off-by: Evgeny Pinchuk
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/kfd_chardev.c | 61
This patch fixes a bug in the code flow that caused an override of the sh_mem
registers.
The bug resulted in sh_mem registers being not initialized properly and
overwrite over sh_mem registers for vmid 0 (which is a vmid of non-HSA
processes).
Reviewed-by: Ben Goz
Signed-off-by: Oded Gabbay
---
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/kfd_chardev.c | 11 ++-
1 file changed, 6 insertions(+), 5 deletions(-)
diff --git a/drivers/gpu/hsa/radeon/kfd_chardev.c
b/drivers/gpu/hsa/radeon/kfd_chardev.c
index 51f790f..09c9a61 100644
--- a/drivers/gpu/hsa/radeon/kfd_chardev.c
From: Yair Shachar
Waiting on fence returns status
Signed-off-by: Yair Shachar
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/kfd_device_queue_manager.c | 6 --
drivers/gpu/hsa/radeon/kfd_priv.h | 2 ++
2 files changed, 6 insertions(+), 2 deletions(-)
diff --git a/
This patch fixes a bug when using the mode of CP hardware
scheduling without oversubscription.
The bug was that the oversubscription check was performed
_after_ the current runlist was destroyed, which caused
the current HSA application to stop working.
This patch moves the oversubscription check
From: Ben Goz
This patch extends the kfd2kgd interface by adding functions
that perform direct register access.
These functions can be called from kfd and will allow to
eliminate all direct register accesses from within the kfd.
Signed-off-by: Ben Goz
Signed-off-by: Oded Gabbay
---
drivers/g
This patch set implements a Heterogeneous System Architecture (HSA) driver
for radeon-family GPUs.
HSA allows different processor types (CPUs, DSPs, GPUs, etc..) to share
system resources more effectively via HW features including shared pageable
memory, userspace-accessible work queues, and p
This patch enables the KFD to retrieve the kfd_process
object from the process's mm_struct. This is needed because kfd_process
lifespan is bound to the process's mm_struct lifespan.
When KFD is notified about an mm_struct tear-down, it checks if the
kfd_process pointer is valid. If so, it releases
To support HSA on KV, we need to limit the number of vmids and pipes
that are available for radeon's use with KV.
This patch reserves VMIDs 8-15 for KFD (so radeon can only use VMIDs
0-7) and also makes radeon thinks that KV has only a single MEC with a single
pipe in it
Signed-off-by: Oded Gabba
This patch adds a new interface to kfd2kgd_calls structure so that
the kfd driver could get the virtual ram size of a specific
radeon device.
Signed-off-by: Oded Gabbay
---
drivers/gpu/drm/radeon/radeon_kfd.c | 12
include/linux/radeon_kfd.h | 1 +
2 files changed, 13 inse
This patch adds the kfd mmap handler that maps the physical address
of a doorbell page to a user-space virtual address. That virtual address
belongs to the process that uses the doorbell page.
This mmap handler is called only from within the kernel and not to be
called from user-mode mmap of /dev/
This patch adds 2 new IOCTL to kfd driver.
The first IOCTL is KFD_IOC_CREATE_QUEUE that is used by the user-mode
application to create a compute queue on the GPU.
The second IOCTL is KFD_IOC_DESTROY_QUEUE that is used by the
user-mode application to destroy an existing compute queue on the GPU.
Update MAINTAINERS and CREDITS files with kfd driver information
Signed-off-by: Oded Gabbay
---
CREDITS | 7 +++
MAINTAINERS | 8
2 files changed, 15 insertions(+)
diff --git a/CREDITS b/CREDITS
index 03343bf..c5f0aeae 100644
--- a/CREDITS
+++ b/CREDITS
@@ -1197,6 +1197,13 @@ S
This patch adds the interrupt handling module, in kfd_interrupt.c,
and its related members in different data structures to the KFD
driver.
The KFD interrupt module maintains an internal interrupt ring per kfd
device. The internal interrupt ring contains interrupts that needs further
handling.The e
This patch enables the use of interrupts in the KFD scheduler when the
scheduler performs its initialization.
It also disables the interrupts when the scheduler stops its work.
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/kfd_sched_cik_static.c | 28 +++
1 file
This patch adds a new callback function to the kgd2kfd interface. The
new callback is for propagating interrupts from radeon driver to the kfd
driver.
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/kfd_module.c | 1 +
include/linux/radeon_kfd.h | 1 +
2 files changed, 2 insertion
This patch adds two new interfaces to the kgd2kfd structure. Those
interfaces are for doing suspend and resume of a kfd device, when its
matching radeon device does suspend and resume.
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/Makefile | 2 +-
drivers/gpu/hsa/radeon/kfd_module.c
The radeon driver can suspend and resume its device. For each device it
suspends/resumes, it should inform the kfd about it, so the kfd could
perform relevant actions per that device.
This patch adds the calls to kfd's suspend and resume functions. The
device is passed as an argument.
Signed-off-
From: Andrew Lewycky
Permanently bind the process to the device.
The binding survives even when all queues are destroyed.
Process exit and device removal terminate the binding.
Signed-off-by: Andrew Lewycky
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/kfd_chardev.c | 27 +++--
From: Andrew Lewycky
This patch adds support in KFD for the hsaKmtSetMemoryPolicy
HSA thunk API call
Signed-off-by: Andrew Lewycky
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/cik_regs.h | 1 +
drivers/gpu/hsa/radeon/kfd_chardev.c | 59 +
drivers
From: Andrew Lewycky
This patch changes the location of the mmu_notifier_invalidate_page function
call inside try_to_unmap_one. The mmu_notifier_invalidate_page function
call tells the IOMMU that a pgae should be invalidated.
The location is changed from after releasing the physical page to
befo
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/kfd_device.c | 28 +++-
1 file changed, 23 insertions(+), 5 deletions(-)
diff --git a/drivers/gpu/hsa/radeon/kfd_device.c
b/drivers/gpu/hsa/radeon/kfd_device.c
index b627e57..a21c095 100644
--- a/drivers/gpu/hsa/radeo
From: Evgeny Pinchuk
Adding API for KFD to be able to query the GPU clock counter.
Signed-off-by: Evgeny Pinchuk
Signed-off-by: Oded Gabbay
---
drivers/gpu/drm/radeon/radeon_kfd.c | 9 +
include/linux/radeon_kfd.h | 1 +
2 files changed, 10 insertions(+)
diff --git a/drivers
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/cik_int.h | 20 ++--
1 file changed, 10 insertions(+), 10 deletions(-)
diff --git a/drivers/gpu/hsa/radeon/cik_int.h b/drivers/gpu/hsa/radeon/cik_int.h
index e98551d..350f0c2 100644
--- a/drivers/gpu/hsa/radeon/cik_int.h
+++ b
From: Evgeny Pinchuk
Implementing a lock for selecting and accessing shader engines and arrays.
This lock will make sure that drm/radeon and hsa/radeon are not colliding when
accessing shader engines and arrays with GRBM_GFX_INDEX register.
Signed-off-by: Evgeny Pinchuk
Signed-off-by: Oded Gabb
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/kfd_sched_cik_static.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/gpu/hsa/radeon/kfd_sched_cik_static.c
b/drivers/gpu/hsa/radeon/kfd_sched_cik_static.c
index 5bfde5c..7573d25 100644
--- a/drivers/gpu/hsa/radeo
This patch creates a workaround for a bug in amd_iommu driver, where the driver
doesn't save all necessary information when going to suspend.
The workaround removes a device from the IOMMU device list on suspend and
register a resumed device in the IOMMU device list.
Signed-off-by: Oded Gabbay
-
From: Evgeny Pinchuk
Changing the source of the max engine clock value.
Signed-off-by: Evgeny Pinchuk
Signed-off-by: Oded Gabbay
---
drivers/gpu/drm/radeon/radeon_kfd.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/gpu/drm/radeon/radeon_kfd.c
b/drivers/gpu/drm/r
From: Alexey Skidanov
Initializing compat_ioctl properly. All ioctls args are packed.
Signed-off-by: Alexey Skidanov
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/kfd_chardev.c | 7 +--
drivers/gpu/hsa/radeon/kfd_priv.h| 4
include/uapi/linux/kfd_ioctl.h | 2 +-
3 f
From: Alexey Skidanov
Signed-off-by: Alexey Skidanov
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/kfd_sched_cik_static.c | 6 --
1 file changed, 4 insertions(+), 2 deletions(-)
diff --git a/drivers/gpu/hsa/radeon/kfd_sched_cik_static.c
b/drivers/gpu/hsa/radeon/kfd_sched_cik_stat
From: Alexey Skidanov
Added apertures initialization and appropriate ioctl
Signed-off-by: Alexey Skidanov
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/Makefile | 2 +-
drivers/gpu/hsa/radeon/kfd_aperture.c | 124 ++
drivers/gpu/hsa/rade
From: Michael Varga
Added debug print statements so critical errors during init may be debugged
more easily.
Signed-off-by: Michael Varga
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/kfd_device.c | 18 +++---
1 file changed, 15 insertions(+), 3 deletions(-)
diff --git a
From: Ben Goz
This patch adds infrastructure to allocate doorbells which are not exposed to
user space.
Signed-off-by: Ben Goz
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/kfd_doorbell.c | 76 ++-
drivers/gpu/hsa/radeon/kfd_priv.h | 5 +++
2 files
From: Ben Goz
The mqd_manager module handles MQD data structures. MQD stands for Memory Queue
Descriptor, which is used by the H/W to keep the HSA queue state in memory.
Signed-off-by: Ben Goz
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/Makefile | 2 +-
drivers/gpu/h
From: Ben Goz
The queue scheduler divides into two sections, one section is process bounded
and the other section is device bounded.
The process bounded section is handled by this module. The PQM handles HSA
queue setup, updates and tear-down.
Signed-off-by: Ben Goz
Signed-off-by: Oded Gabbay
From: Ben Goz
The queue scheduler divides into two sections, one section is process bounded
and the other section is device bounded.
The device bounded section is handled by this module.
The DQM module handles queue setup, update and tear-down from the device side.
It also supports suspend/resume
From: Ben Goz
This patch makes the switch between the old KFD queue scheduler to the new KFD
queue scheduler. The new scheduler supports H/W CP scheduling, over-subscription
of queues and pre-emption of queues.
Signed-off-by: Ben Goz
Signed-off-by: Oded Gabbay
---
drivers/gpu/hsa/radeon/kfd_a
1 - 100 of 185 matches
Mail list logo
