> improvements made in [2]. Despite this, I have proceeded to submit this
> patch because of the noticeable performance improvements it provides.
>
> Reported-by: Vitaly Kuznetsov
FWIW, this patch (alone) resolves the observed issue, thanks!
Tested-by: Vitaly Kuznetsov
I, however, sha
eahar...@linux.microsoft.com writes:
> From: Easwar Hariharan
>
> Move the Microsoft PCI Vendor ID from the various drivers to the pci_ids
> file
>
> Signed-off-by: Easwar Hariharan
> ---
> drivers/gpu/drm/hyperv/hyperv_drm_drv.c | 1 -
> drivers/net/ethernet/microsoft/mana/gdma_main.c
Bjorn Helgaas writes:
> On Sat, Aug 27, 2022 at 03:03:43PM +0200, Vitaly Kuznetsov wrote:
>> There are already three places in kernel which define PCI_VENDOR_ID_MICROSOFT
>> and two for PCI_DEVICE_ID_HYPERV_VIDEO and there's a need to use these
>> from core Vmbus co
g besides framebuffer from there (PATCH3).
Vitaly Kuznetsov (3):
PCI: Move PCI_VENDOR_ID_MICROSOFT/PCI_DEVICE_ID_HYPERV_VIDEO
definitions to pci_ids.h
Drivers: hv: Always reserve framebuffer region for Gen1 VMs
Drivers: hv: Never allocate anything besides framebuffer from
framebuffe
e this region unclaimed by the time Hyper-V PCI
pass-through driver tries allocating MMIO space if Hyper-V DRM/FB drivers
load after it. Devices can be brought up in any sequence so let's
resolve the issue by always ignoring 'fb_mmio' region for non-FB
requests, even if the region
ids.h
Signed-off-by: Vitaly Kuznetsov
---
drivers/gpu/drm/hyperv/hyperv_drm_drv.c | 3 ---
drivers/net/ethernet/microsoft/mana/gdma_main.c | 4
drivers/video/fbdev/hyperv_fb.c | 4
include/linux/pci_ids.h | 3 +++
4 files changed, 3 inser
ould always be
possible to get the address from the dedicated PCI device instead.
Check for legacy PCI video device presence and reserve the whole
region for framebuffer on Gen1 VMs.
Reviewed-by: Michael Kelley
Signed-off-by: Vitaly Kuznetsov
---
drivers/hv/vmbus_drv.c | 46
"Michael Kelley (LINUX)" writes:
> From: Vitaly Kuznetsov Sent: Thursday, August 25, 2022
> 2:00 AM
>>
>> Passed through PCI device sometimes misbehave on Gen1 VMs when Hyper-V
>> DRM driver is also loaded. Looking at IOMEM assignment, we can
"Michael Kelley (LINUX)" writes:
> From: Vitaly Kuznetsov Sent: Thursday, August 25, 2022
> 2:00 AM
>>
>> vmbus_reserve_fb() tries reserving framebuffer region iff
>> 'screen_info.lfb_base' is set. Gen2 VMs seem to have it set by EFI fb
>
&g
There are already three places in kernel which define PCI_VENDOR_ID_MICROSOFT
and two for PCI_DEVICE_ID_HYPERV_VIDEO and there's a need to use these
from core Vmbus code. Move the defines where they belong.
No functional change.
Signed-off-by: Vitaly Kuznetsov
---
drivers/gpu/drm/h
device which is always present.
Check for legacy PCI video device presence and reserve the whole
region for framebuffer on Gen1 VMs.
Signed-off-by: Vitaly Kuznetsov
---
drivers/hv/vmbus_drv.c | 46 +-
1 file changed, 32 insertions(+), 14 deletions(-)
diff --
always ignoring 'fb_mmio' region for non-FB
requests, even if the region is unclaimed.
Signed-off-by: Vitaly Kuznetsov
---
drivers/hv/vmbus_drv.c | 10 +-
1 file changed, 9 insertions(+), 1 deletion(-)
diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c
index
FB region on Gen1 VMs (PATCH2) and make sure we never allocate
anything besides framebuffer from there (PATCH3).
Vitaly Kuznetsov (3):
PCI: Move PCI_VENDOR_ID_MICROSOFT/PCI_DEVICE_ID_HYPERV_VIDEO
definitions to pci_ids.h
Drivers: hv: Always reserve framebuffer region for Gen1 VMs
Drivers:
"Michael Kelley (LINUX)" writes:
> From: Vitaly Kuznetsov Sent: Thursday, August 18, 2022
> 7:25 AM
>>
>> When drm_aperture_remove_conflicting_pci_framebuffers() fails, 'pdev'
>> needs to be released with pci_dev_put().
>>
>> Fix
When drm_aperture_remove_conflicting_pci_framebuffers() fails, 'pdev'
needs to be released with pci_dev_put().
Fixes: 76c56a5affeb ("drm/hyperv: Add DRM driver for hyperv synthetic video
device")
Signed-off-by: Vitaly Kuznetsov
---
drivers/gpu/drm/hyperv/hyperv_drm_drv.c |
"Michael Kelley (LINUX)" writes:
> From: Vitaly Kuznetsov Sent: Thursday, August 18, 2022
> 7:25 AM
>>
>> There are already two places in kernel with PCI_VENDOR_ID_MICROSOFT/
>> PCI_DEVICE_ID_HYPERV_VIDEO and there's a need to use these from core
>
FB region on Gen1 VMs (PATCH3) and making
sure we never allocate anything besides framebuffer from there (PATCH4). PATCH1
is a preparatory change, PATCH2 fixes a loosely related issue in Hyper-V DRM
driver.
Vitaly Kuznetsov (4):
Drivers: hv: Move legacy Hyper-V PCI video device's ids to
linu
. Both Hyper-V DRM and Hyper-V FB
drivers get framebuffer base from this PCI device already so Vmbus
driver can do the same trick.
Check for legacy PCI video device presence and reserve the whole
region for framebuffer.
Signed-off-by: Vitaly Kuznetsov
---
drivers/hv/vmbus_drv.c | 47
There are already two places in kernel with PCI_VENDOR_ID_MICROSOFT/
PCI_DEVICE_ID_HYPERV_VIDEO and there's a need to use these from core
Vmbus code. Move the defines to a common header.
No functional change.
Signed-off-by: Vitaly Kuznetsov
---
drivers/gpu/drm/hyperv/hyperv_drm_drv.
always ignoring 'fb_mmio' region for non-FB
requests, even if the region is unclaimed.
Signed-off-by: Vitaly Kuznetsov
---
drivers/hv/vmbus_drv.c | 10 +-
1 file changed, 9 insertions(+), 1 deletion(-)
diff --git a/drivers/hv/vmbus_drv.c b/drivers/hv/vmbus_drv.c
index
Hi,
Commit ccf953d8f3d6 ("fb_defio: Remove custom address_space_operations")
seems to be breaking Hyper-V framebuffer
(drivers/video/fbdev/hyperv_fb.c) driver for me: Hyper-V guest boots
well and plymouth even works but when I try starting Gnome, virtual
screen just goes black. Reverting the above
Deepak Rawat writes:
> On Wed, 2021-02-17 at 13:07 +0100, Vitaly Kuznetsov wrote:
>> > +++ b/drivers/gpu/drm/hyperv/hyperv_drm.h
>> > @@ -0,0 +1,51 @@
>> > +/* SPDX-License-Identifier: GPL-2.0 */
>> > +/*
>> > + * Copyright 2012-2021 Microsoft
>
Deepak Rawat writes:
> DRM driver for hyperv synthetic video device, based on hyperv_fb
> framebuffer driver. Also added config option "DRM_HYPERV" to enabled
> this driver.
>
> v2:
> - Add support for gen2 VM
> - Fixed review comments
>
> v3:
> - Split into multiple files as suggested by Thomas
23 matches
Mail list logo