similar to: RFC: xl and HVM usb devices

Ian Campbell wrote: > On Mon, 2013-09-09 at 03:29 +0000, Liu, Jinsong wrote: >> From 18344216b432648605726b137b348f28ef64a4ef Mon Sep 17 00:00:00 >> 2001 >> From: Liu Jinsong <jinsong.liu@intel.com> >> Date: Fri, 23 Aug 2013 23:30:23 +0800 >> Subject: [PATCH V3] xl: HVM domain S3 bugfix >> >> Currently Xen hvm s3 has a bug coming from the

From 18344216b432648605726b137b348f28ef64a4ef Mon Sep 17 00:00:00 2001 From: Liu Jinsong <jinsong.liu@intel.com> Date: Fri, 23 Aug 2013 23:30:23 +0800 Subject: [PATCH V3] xl: HVM domain S3 bugfix Currently Xen hvm s3 has a bug coming from the difference between qemu-traditioanl and qemu-xen. For qemu-traditional, the way to resume from hvm s3 is via ''xl trigger'' command.

This patch exposes a generic interface which can be expanded in the future to implement USB for PVUSB, qemu, and stubdoms. It can also be extended to include other types of USB other than host USB (for example, tablets, mice, or keyboards). For each device removed or added, one of two protocols is available: * PVUSB * qemu (DEVICEMODEL) The caller can additionally specify "AUTO", in

According to the xen 4.1 release notes linked from here: http://xen.org/products/xen_source.html Some features are missing from xen 4.1. i.e. *List of missing features from xl/libxl in Xen 4.1:* ** - *PVUSB (for more info see PVUSB sections of: XenUSBPassthrough<http://wiki.xen.org/xenwiki/XenUSBPassthrough.html> )* - *PVSCSI (for more info see:

Upcasting using a container_of macro is more typesafe, faster and easier for the compiler to optimize. Acked-by: Gerd Hoffmann <kraxel at redhat.com> Signed-off-by: Daniel Vetter <daniel.vetter at intel.com> Cc: Dave Airlie <airlied at redhat.com> Cc: Gerd Hoffmann <kraxel at redhat.com> Cc: virtualization at lists.linux-foundation.org Cc: spice-devel at

Move final cleanups to qxl_drm_release() callback. Add drm_atomic_helper_shutdown() call to qxl_pci_remove(). Reorder calls in qxl_device_fini(). Cleaning up gem & ttm might trigger qxl commands, so we should do that before releaseing command rings. Signed-off-by: Gerd Hoffmann <kraxel at redhat.com> --- drivers/gpu/drm/qxl/qxl_drv.c | 21 ++++++++++++++-------

Also need to remove the drm_dev_put from the remove hook. Acked-by: Gerd Hoffmann <kraxel at redhat.com> Signed-off-by: Daniel Vetter <daniel.vetter at intel.com> Cc: Dave Airlie <airlied at redhat.com> Cc: Gerd Hoffmann <kraxel at redhat.com> Cc: virtualization at lists.linux-foundation.org Cc: spice-devel at lists.freedesktop.org --- drivers/gpu/drm/qxl/qxl_drv.c | 15

Also need to remove the drm_dev_put from the remove hook. Acked-by: Gerd Hoffmann <kraxel at redhat.com> Signed-off-by: Daniel Vetter <daniel.vetter at intel.com> Cc: Dave Airlie <airlied at redhat.com> Cc: Gerd Hoffmann <kraxel at redhat.com> Cc: virtualization at lists.linux-foundation.org Cc: spice-devel at lists.freedesktop.org --- drivers/gpu/drm/qxl/qxl_drv.c | 15

Also need to remove the drm_dev_put from the remove hook. Signed-off-by: Daniel Vetter <daniel.vetter at intel.com> Cc: Dave Airlie <airlied at redhat.com> Cc: Gerd Hoffmann <kraxel at redhat.com> Cc: virtualization at lists.linux-foundation.org Cc: spice-devel at lists.freedesktop.org --- drivers/gpu/drm/qxl/qxl_drv.c | 15 ++++++++------- drivers/gpu/drm/qxl/qxl_drv.h | 3

Also need to remove the drm_dev_put from the remove hook. Signed-off-by: Daniel Vetter <daniel.vetter at intel.com> Cc: Dave Airlie <airlied at redhat.com> Cc: Gerd Hoffmann <kraxel at redhat.com> Cc: virtualization at lists.linux-foundation.org Cc: spice-devel at lists.freedesktop.org --- drivers/gpu/drm/qxl/qxl_drv.c | 15 ++++++++------- drivers/gpu/drm/qxl/qxl_drv.h | 3

Move final cleanups to qxl_drm_release() callback. Add drm_atomic_helper_shutdown() call to qxl_pci_remove(). Reorder calls in qxl_device_fini(). Cleaning up gem & ttm might trigger qxl commands, so we should do that before releaseing command rings. Signed-off-by: Gerd Hoffmann <kraxel at redhat.com> --- drivers/gpu/drm/qxl/qxl_drv.c | 21 ++++++++++++++-------

On Fri, Apr 24, 2020 at 05:09:11PM +0200, Sam Ravnborg wrote: > Hi Daniel > > On Wed, Apr 15, 2020 at 09:40:01AM +0200, Daniel Vetter wrote: > > Also need to remove the drm_dev_put from the remove hook. > > > > Acked-by: Gerd Hoffmann <kraxel at redhat.com> > > Signed-off-by: Daniel Vetter <daniel.vetter at intel.com> > > Cc: Dave Airlie

On Fri, Apr 24, 2020 at 05:09:11PM +0200, Sam Ravnborg wrote: > Hi Daniel > > On Wed, Apr 15, 2020 at 09:40:01AM +0200, Daniel Vetter wrote: > > Also need to remove the drm_dev_put from the remove hook. > > > > Acked-by: Gerd Hoffmann <kraxel at redhat.com> > > Signed-off-by: Daniel Vetter <daniel.vetter at intel.com> > > Cc: Dave Airlie

With this we can drop the final kfree from the release function. Signed-off-by: Daniel Vetter <daniel.vetter at intel.com> Cc: Dave Airlie <airlied at redhat.com> Cc: Gerd Hoffmann <kraxel at redhat.com> Cc: virtualization at lists.linux-foundation.org Cc: spice-devel at lists.freedesktop.org --- drivers/gpu/drm/qxl/qxl_drv.c | 2 -- drivers/gpu/drm/qxl/qxl_kms.c | 2 ++ 2

Signed-off-by: Gerd Hoffmann <kraxel at redhat.com> --- drivers/gpu/drm/qxl/qxl_display.c | 5 +++-- 1 file changed, 3 insertions(+), 2 deletions(-) diff --git a/drivers/gpu/drm/qxl/qxl_display.c b/drivers/gpu/drm/qxl/qxl_display.c index fa79688013b7..4be04eaf7f37 100644 --- a/drivers/gpu/drm/qxl/qxl_display.c +++ b/drivers/gpu/drm/qxl/qxl_display.c @@ -1190,7 +1190,9 @@ int

Move final cleanups to qxl_drm_release() callback. Add drm_atomic_helper_shutdown() call to qxl_pci_remove(). Signed-off-by: Gerd Hoffmann <kraxel at redhat.com> --- drivers/gpu/drm/qxl/qxl_drv.c | 26 +++++++++++++++++++------- 1 file changed, 19 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/qxl/qxl_drv.c b/drivers/gpu/drm/qxl/qxl_drv.c index 1d601f57a6ba..4fda3f9b29f4

Move final cleanups to qxl_drm_release() callback. Add drm_atomic_helper_shutdown() call to qxl_pci_remove(). Signed-off-by: Gerd Hoffmann <kraxel at redhat.com> --- drivers/gpu/drm/qxl/qxl_drv.c | 26 +++++++++++++++++++------- 1 file changed, 19 insertions(+), 7 deletions(-) diff --git a/drivers/gpu/drm/qxl/qxl_drv.c b/drivers/gpu/drm/qxl/qxl_drv.c index 1d601f57a6ba..4fda3f9b29f4

s/PAGE_SIZE/PAGE_MASK/ Luckily release_offset is never larger than PAGE_SIZE, so the bug has no bad side effects and managed to stay unnoticed for years that way ... Signed-off-by: Gerd Hoffmann <kraxel at redhat.com> --- drivers/gpu/drm/qxl/qxl_ioctl.c | 4 ++-- drivers/gpu/drm/qxl/qxl_release.c | 6 +++--- 2 files changed, 5 insertions(+), 5 deletions(-) diff --git

Package: xen-system-amd64 Version: 4.17.1+2-gb773c48e36-1 Severity: important Dear Maintainer, after upgrading my bullseye server to bookworm I ran into the issue that all VMs of type hvm are not starting anymore. xl throws an error: libxl: error: libxl_qmp.c:1399:qmp_ev_fd_callback: Domain 8:error on QMP socket: Connection reset by peer libxl: error: libxl_qmp.c:1438:qmp_ev_fd_callback: Domain

qxl has two modes: "native" (used by the drm driver) and "vga" (vga compatibility mode, typically used for boot display and firmware framebuffers). Accessing any vga ioport will switch the qxl device into vga mode. The qxl driver never does that, but other drivers accessing vga ports can trigger that too and therefore disturb qxl operation. So aquire the legacy vga ioports