Skip to contentSkip to conversation

libxl: try to support IGD passthrough for qemu upstream

PATCH In review Tiejun Chen 3 participants 10 messages Sep 18, 2015 – Oct 6, 2015

lore
Share & tools
lore
#0 lore raw

v4 cover libxl: try to support IGD passthrough for qemu upstream

Ian,

As we discussed previously,

http://patchwork.ozlabs.org/patch/457055/

now it's time to push this into on xen/tools side since all qemu stuffs have been merged.

https://lists.gnu.org/archive/html/qemu-devel/2015-09/msg02094.html

v4:

Ian,

Actually we had v3.5 online previously, which was reviewed by you.

http://permalink.gmane.org/gmane.comp.emulators.qemu/329100

So here I just bring a little bit to refine code just for patch #2 according to out last conversation.

v3:

* Refine some codes based on Campbell's feedback so thanks for Campbell's kind guideline to patch #2
* Update the manpages in patch #2

v2:

* Refine patch #2's head description
* Improve codes quality inside patch #1 based on Wei's comments
* Refill the summary inside patch #0 based on Konrad and Wei's suggestion

When we're working to support IGD GFX passthrough with qemu upstream, instead of "-gfx_passthru" we'd like to make that a machine option, "-machine xxx,igd-passthru=on".

https://lists.nongnu.org/archive/html/qemu-devel/2015-01/msg02050.html

This need to bring a change on tool side.

After a discussion with Campbell, we'd like to construct a table to record all IGD devices we can support. If we hit that table, we should pass that option. And so we also introduce a new field of type, 'gfx_passthru_kind', to cooperate with 'gfx_passthru' to cover all scenarios like this,

gfx_passthru = 0 => sets build_info.u.gfx_passthru to false
gfx_passthru = 1 => sets build_info.u.gfx_passthru to true and
build_info.u.gfx_passthru_kind to DEFAULT
gfx_passthru = "igd" => sets build_info.u.gfx_passthru to false
and build_info.u.gfx_passthru_kind to IGD

And note actually that option "-gfx_passthru" is just introduced to work for qemu-xen-traditional so we should get this away from libxl__build_device_model_args_new() in the case of qemu upstream.

----------------------------------------------------------------
Tiejun Chen (2):
libxl: introduce libxl__is_igd_vga_passthru
libxl: introduce gfx_passthru_kind

 docs/man/xl.cfg.pod.5        |  35 ++++++++--
 tools/libxl/libxl.h          |   6 ++
 tools/libxl/libxl_dm.c       |  46 +++++++++++--
 tools/libxl/libxl_internal.h |   2 +
 tools/libxl/libxl_pci.c      | 124 +++++++++++++++++++++++++++++++++++
 tools/libxl/libxl_types.idl  |   6 ++
 tools/libxl/xl_cmdimpl.c     |  14 +++-
 7 files changed, 223 insertions(+), 10 deletions(-)

Thanks
Tiejun

#1 lore raw

PATCH v4 1/2 libxl: introduce libxl__is_igd_vga_passthru

While working with qemu, IGD is a specific device in the case of pass through so we need to identify that to handle more later. Here we define a table to record all IGD types currently we can support. Also we need to introduce two helper functions to get vendor and device ids to lookup that table.

---
 tools/libxl/libxl_internal.h |   2 +
 tools/libxl/libxl_pci.c      | 124 +++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 126 insertions(+)
tools/libxl/libxl_internal.h +2 -0
index 5fa55a7..5d6c793 100644
@@ -1343,6 +1343,8 @@ _hidden int libxl__device_pci_add(libxl__gc *gc, uint32_t domid, libxl_device_pc
 _hidden int libxl__create_pci_backend(libxl__gc *gc, uint32_t domid,
                                       libxl_device_pci *pcidev, int num);
 _hidden int libxl__device_pci_destroy_all(libxl__gc *gc, uint32_t domid);
+_hidden bool libxl__is_igd_vga_passthru(libxl__gc *gc,
+                                        const libxl_domain_config *d_config);
 
 /* from libxl_dtdev */
 
tools/libxl/libxl_pci.c +124 -0

130 lines not on this page · raw message

Signature · 1 line
1.9.1
#2 lore raw

PATCH v4 1/2 libxl: introduce libxl__is_igd_vga_passthru

In reply to Tiejun Chen

While working with qemu, IGD is a specific device in the case of pass through so we need to identify that to handle more later. Here we define a table to record all IGD types currently we can support. Also we need to introduce two helper functions to get vendor and device ids to lookup that table.

---
 tools/libxl/libxl_internal.h |   2 +
 tools/libxl/libxl_pci.c      | 124 +++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 126 insertions(+)
tools/libxl/libxl_internal.h +2 -0
index 5fa55a7..5d6c793 100644
@@ -1343,6 +1343,8 @@ _hidden int libxl__device_pci_add(libxl__gc *gc, uint32_t domid, libxl_device_pc
 _hidden int libxl__create_pci_backend(libxl__gc *gc, uint32_t domid,
                                       libxl_device_pci *pcidev, int num);
 _hidden int libxl__device_pci_destroy_all(libxl__gc *gc, uint32_t domid);
+_hidden bool libxl__is_igd_vga_passthru(libxl__gc *gc,
+                                        const libxl_domain_config *d_config);
 
 /* from libxl_dtdev */
 
tools/libxl/libxl_pci.c +124 -0

130 lines not on this page · raw message

Signature · 1 line
1.9.1
#3 lore raw

PATCH v4 2/2 libxl: introduce gfx_passthru_kind

In reply to Tiejun Chen

Although we already have 'gfx_passthru' in b_info, this doesn't suffice after we want to handle IGD specifically. Now we define a new field of type, gfx_passthru_kind, to indicate we're trying to pass IGD. Actually this means we can benefit this to support other specific devices just by extending gfx_passthru_kind. And then we can cooperate with gfx_passthru to address IGD cases as follows:

gfx_passthru = 0 => sets build_info.u.gfx_passthru to false
gfx_passthru = 1 => sets build_info.u.gfx_passthru to true and
build_info.u.gfx_passthru_kind to DEFAULT
gfx_passthru = "igd" => sets build_info.u.gfx_passthru to true
and build_info.u.gfx_passthru_kind to IGD

Here if gfx_passthru_kind = DEFAULT, we will call libxl__is_igd_vga_passthru() to check if we're hitting that table to need to pass that option to qemu. But if gfx_passthru_kind = "igd" we always force to pass that.

And "-gfx_passthru" is just introduced to work for qemu-xen-traditional so we should get this away from libxl__build_device_model_args_new() in the case of qemu upstream.

---
 docs/man/xl.cfg.pod.5       | 35 ++++++++++++++++++++++++++++++----
 tools/libxl/libxl.h         |  6 ++++++
 tools/libxl/libxl_dm.c      | 46 +++++++++++++++++++++++++++++++++++++++++----
 tools/libxl/libxl_types.idl |  6 ++++++
 tools/libxl/xl_cmdimpl.c    | 14 ++++++++++++--
 5 files changed, 97 insertions(+), 10 deletions(-)
docs/man/xl.cfg.pod.5 +31 -4
index db4a163..4aa7b05 100644
@@ -812,7 +812,7 @@ through to this VM. See L<seize|/"seize_boolean"> above.
 devices passed through to this VM. See L<power_mgt|/"power_mgmt_boolean">
 above.
 
-=item B<gfx_passthru=BOOLEAN>
+=item B<gfx_passthru=BOOLEAN|"STRING">
 
 Enable graphics device PCI passthrough. This option makes an assigned
 PCI graphics card become primary graphics card in the VM. The QEMU
@@ -840,9 +840,36 @@ working graphics passthrough. See the XenVGAPassthroughTestedAdapters
 L<http://wiki.xen.org/wiki/XenVGAPassthroughTestedAdapters> wiki page
 for currently supported graphics cards for gfx_passthru.
 
-gfx_passthru is currently only supported with the qemu-xen-traditional
-device-model. Upstream qemu-xen device-model currently does not have
-support for gfx_passthru.
+gfx_passthru is currently supported both with the qemu-xen-traditional
+device-model and upstream qemu-xen device-model.
+
+When given as a boolean the B<gfx_passthru> option either disables gfx
+passthru or enables autodetection.
+
+But when given as a string the B<gfx_passthru> option describes the type
+of device to enable. Note this behavior is only supported with the upstream
+qemu-xen device-model. With qemu-xen-traditional IGD is always assumed
+and other options than autodetect or explicit IGD will result in an error.
+
+Currently, valid options are:
+
+=over 4
+
+=item B<gfx_passthru=0>
+
+Disables graphics device PCI passthrough.
+
+=item B<gfx_passthru=1>, B<gfx_passthru="default">
+
+Enables graphics device PCI passthrough and autodetects the type of device
+which is being used.
+
+=item "igd"
+
+Enables graphics device PCI passthrough but forcing the type of device to
+Intel Graphics Device.
+
+=back
 
 Note that some graphics adapters (AMD/ATI cards, for example) do not
 necessarily require gfx_passthru option, so you can use the normal Xen
tools/libxl/libxl.h +6 -0
index 5f9047c..899de28 100644
@@ -832,6 +832,12 @@ void libxl_mac_copy(libxl_ctx *ctx, libxl_mac *dst, libxl_mac *src);
  */
 #define LIBXL_HAVE_SRM_V1 1
 
+/*
+ * libxl_domain_build_info has the u.hvm.gfx_passthru_kind field and
+ * the libxl_gfx_passthru_kind enumeration is defined.
+*/
+#define LIBXL_HAVE_GFX_PASSTHRU_KIND
+
 typedef char **libxl_string_list;
 void libxl_string_list_dispose(libxl_string_list *sl);
 int libxl_string_list_length(const libxl_string_list *sl);
tools/libxl/libxl_dm.c +42 -4

70 lines not on this page · raw message

tools/libxl/libxl_types.idl +6 -0
index 9f6ec00..7e1ab85 100644
@@ -157,6 +157,11 @@ libxl_tsc_mode = Enumeration("tsc_mode", [
     (3, "native_paravirt"),
     ])
 
+libxl_gfx_passthru_kind = Enumeration("gfx_passthru_kind", [
+    (0, "default"),
+    (1, "igd"),
+    ])
+
 # Consistent with the values defined for HVM_PARAM_TIMER_MODE.
 libxl_timer_mode = Enumeration("timer_mode", [
     (-1, "unknown"),
@@ -488,6 +493,7 @@ libxl_domain_build_info = Struct("domain_build_info",[
                                        ("spice",            libxl_spice_info),
                                        
                                        ("gfx_passthru",     libxl_defbool),
+                                       ("gfx_passthru_kind", libxl_gfx_passthru_kind),
                                        
                                        ("serial",           string),
                                        ("boot",             string),
tools/libxl/xl_cmdimpl.c +12 -2
index 2706759..b26fc37 100644
@@ -2295,8 +2295,18 @@ skip_vfb:
         xlu_cfg_replace_string (config, "spice_streaming_video",
                                 &b_info->u.hvm.spice.streaming_video, 0);
         xlu_cfg_get_defbool(config, "nographic", &b_info->u.hvm.nographic, 0);
-        xlu_cfg_get_defbool(config, "gfx_passthru",
-                            &b_info->u.hvm.gfx_passthru, 0);
+        if (!xlu_cfg_get_long(config, "gfx_passthru", &l, 1)) {
+            libxl_defbool_set(&b_info->u.hvm.gfx_passthru, l);
+        } else if (!xlu_cfg_get_string(config, "gfx_passthru", &buf, 0)) {
+            if (libxl_gfx_passthru_kind_from_string(buf,
+                                        &b_info->u.hvm.gfx_passthru_kind)) {
+                fprintf(stderr,
+                        "ERROR: invalid value \"%s\" for \"gfx_passthru\"\n",
+                        buf);
+                exit (1);
+            }
+            libxl_defbool_set(&b_info->u.hvm.gfx_passthru, true);
+        }
         switch (xlu_cfg_get_list_as_string_list(config, "serial",
                                                 &b_info->u.hvm.serial_list,
                                                 1))
Signature · 1 line
1.9.1
#4 lore raw

PATCH v4 2/2 libxl: introduce gfx_passthru_kind

In reply to Tiejun Chen

Although we already have 'gfx_passthru' in b_info, this doesn't suffice after we want to handle IGD specifically. Now we define a new field of type, gfx_passthru_kind, to indicate we're trying to pass IGD. Actually this means we can benefit this to support other specific devices just by extending gfx_passthru_kind. And then we can cooperate with gfx_passthru to address IGD cases as follows:

gfx_passthru = 0 => sets build_info.u.gfx_passthru to false
gfx_passthru = 1 => sets build_info.u.gfx_passthru to true and
build_info.u.gfx_passthru_kind to DEFAULT
gfx_passthru = "igd" => sets build_info.u.gfx_passthru to true
and build_info.u.gfx_passthru_kind to IGD

Here if gfx_passthru_kind = DEFAULT, we will call libxl__is_igd_vga_passthru() to check if we're hitting that table to need to pass that option to qemu. But if gfx_passthru_kind = "igd" we always force to pass that.

And "-gfx_passthru" is just introduced to work for qemu-xen-traditional so we should get this away from libxl__build_device_model_args_new() in the case of qemu upstream.

---
 docs/man/xl.cfg.pod.5       | 35 ++++++++++++++++++++++++++++++----
 tools/libxl/libxl.h         |  6 ++++++
 tools/libxl/libxl_dm.c      | 46 +++++++++++++++++++++++++++++++++++++++++----
 tools/libxl/libxl_types.idl |  6 ++++++
 tools/libxl/xl_cmdimpl.c    | 14 ++++++++++++--
 5 files changed, 97 insertions(+), 10 deletions(-)
docs/man/xl.cfg.pod.5 +31 -4
index db4a163..4aa7b05 100644
@@ -812,7 +812,7 @@ through to this VM. See L<seize|/"seize_boolean"> above.
 devices passed through to this VM. See L<power_mgt|/"power_mgmt_boolean">
 above.
 
-=item B<gfx_passthru=BOOLEAN>
+=item B<gfx_passthru=BOOLEAN|"STRING">
 
 Enable graphics device PCI passthrough. This option makes an assigned
 PCI graphics card become primary graphics card in the VM. The QEMU
@@ -840,9 +840,36 @@ working graphics passthrough. See the XenVGAPassthroughTestedAdapters
 L<http://wiki.xen.org/wiki/XenVGAPassthroughTestedAdapters> wiki page
 for currently supported graphics cards for gfx_passthru.
 
-gfx_passthru is currently only supported with the qemu-xen-traditional
-device-model. Upstream qemu-xen device-model currently does not have
-support for gfx_passthru.
+gfx_passthru is currently supported both with the qemu-xen-traditional
+device-model and upstream qemu-xen device-model.
+
+When given as a boolean the B<gfx_passthru> option either disables gfx
+passthru or enables autodetection.
+
+But when given as a string the B<gfx_passthru> option describes the type
+of device to enable. Note this behavior is only supported with the upstream
+qemu-xen device-model. With qemu-xen-traditional IGD is always assumed
+and other options than autodetect or explicit IGD will result in an error.
+
+Currently, valid options are:
+
+=over 4
+
+=item B<gfx_passthru=0>
+
+Disables graphics device PCI passthrough.
+
+=item B<gfx_passthru=1>, B<gfx_passthru="default">
+
+Enables graphics device PCI passthrough and autodetects the type of device
+which is being used.
+
+=item "igd"
+
+Enables graphics device PCI passthrough but forcing the type of device to
+Intel Graphics Device.
+
+=back
 
 Note that some graphics adapters (AMD/ATI cards, for example) do not
 necessarily require gfx_passthru option, so you can use the normal Xen
tools/libxl/libxl.h +6 -0
index 5f9047c..899de28 100644
@@ -832,6 +832,12 @@ void libxl_mac_copy(libxl_ctx *ctx, libxl_mac *dst, libxl_mac *src);
  */
 #define LIBXL_HAVE_SRM_V1 1
 
+/*
+ * libxl_domain_build_info has the u.hvm.gfx_passthru_kind field and
+ * the libxl_gfx_passthru_kind enumeration is defined.
+*/
+#define LIBXL_HAVE_GFX_PASSTHRU_KIND
+
 typedef char **libxl_string_list;
 void libxl_string_list_dispose(libxl_string_list *sl);
 int libxl_string_list_length(const libxl_string_list *sl);
tools/libxl/libxl_dm.c +42 -4

70 lines not on this page · raw message

tools/libxl/libxl_types.idl +6 -0
index 9f6ec00..7e1ab85 100644
@@ -157,6 +157,11 @@ libxl_tsc_mode = Enumeration("tsc_mode", [
     (3, "native_paravirt"),
     ])
 
+libxl_gfx_passthru_kind = Enumeration("gfx_passthru_kind", [
+    (0, "default"),
+    (1, "igd"),
+    ])
+
 # Consistent with the values defined for HVM_PARAM_TIMER_MODE.
 libxl_timer_mode = Enumeration("timer_mode", [
     (-1, "unknown"),
@@ -488,6 +493,7 @@ libxl_domain_build_info = Struct("domain_build_info",[
                                        ("spice",            libxl_spice_info),
                                        
                                        ("gfx_passthru",     libxl_defbool),
+                                       ("gfx_passthru_kind", libxl_gfx_passthru_kind),
                                        
                                        ("serial",           string),
                                        ("boot",             string),
tools/libxl/xl_cmdimpl.c +12 -2
index 2706759..b26fc37 100644
@@ -2295,8 +2295,18 @@ skip_vfb:
         xlu_cfg_replace_string (config, "spice_streaming_video",
                                 &b_info->u.hvm.spice.streaming_video, 0);
         xlu_cfg_get_defbool(config, "nographic", &b_info->u.hvm.nographic, 0);
-        xlu_cfg_get_defbool(config, "gfx_passthru",
-                            &b_info->u.hvm.gfx_passthru, 0);
+        if (!xlu_cfg_get_long(config, "gfx_passthru", &l, 1)) {
+            libxl_defbool_set(&b_info->u.hvm.gfx_passthru, l);
+        } else if (!xlu_cfg_get_string(config, "gfx_passthru", &buf, 0)) {
+            if (libxl_gfx_passthru_kind_from_string(buf,
+                                        &b_info->u.hvm.gfx_passthru_kind)) {
+                fprintf(stderr,
+                        "ERROR: invalid value \"%s\" for \"gfx_passthru\"\n",
+                        buf);
+                exit (1);
+            }
+            libxl_defbool_set(&b_info->u.hvm.gfx_passthru, true);
+        }
         switch (xlu_cfg_get_list_as_string_list(config, "serial",
                                                 &b_info->u.hvm.serial_list,
                                                 1))
Signature · 1 line
1.9.1
#5 lore raw

Re: [v4][PATCH 2/2] libxl: introduce gfx_passthru_kind

Quoted 18 lines · Tiejun Chen

Although we already have 'gfx_passthru' in b_info, this doesn't suffice after we want to handle IGD specifically. Now we define a new field of type, gfx_passthru_kind, to indicate we're trying to pass IGD. Actually this means we can benefit this to support other specific devices just by extending gfx_passthru_kind. And then we can cooperate with gfx_passthru to address IGD cases as follows:

gfx_passthru = 0 => sets build_info.u.gfx_passthru to false
gfx_passthru = 1 => sets build_info.u.gfx_passthru to true and
build_info.u.gfx_passthru_kind to DEFAULT
gfx_passthru = "igd" => sets build_info.u.gfx_passthru to true
and build_info.u.gfx_passthru_kind to IGD

Here if gfx_passthru_kind = DEFAULT, we will call libxl__is_igd_vga_passthru() to check if we're hitting that table to need to pass that option to qemu. But if gfx_passthru_kind = "igd" we always force to pass that.

Quoted 5 lines

And "-gfx_passthru" is just introduced to work for qemu-xen-traditional so we should get this away from libxl__build_device_model_args_new() in the case of qemu upstream.

Acked + applied both patches, thanks.

Stefano -- are the QEMU side patches in qemu-upstream-unstable.git yet? If not I suppose this is a call/reminder to backport them from mainline or whatever.

Ian.

#6 lore raw

Re: [Qemu-devel] [v4][PATCH 2/2] libxl: introduce gfx_passthru_kind

In reply to Tiejun Chen
Quoted 18 lines · Tiejun Chen

Although we already have 'gfx_passthru' in b_info, this doesn't suffice after we want to handle IGD specifically. Now we define a new field of type, gfx_passthru_kind, to indicate we're trying to pass IGD. Actually this means we can benefit this to support other specific devices just by extending gfx_passthru_kind. And then we can cooperate with gfx_passthru to address IGD cases as follows:

gfx_passthru = 0 => sets build_info.u.gfx_passthru to false
gfx_passthru = 1 => sets build_info.u.gfx_passthru to true and
build_info.u.gfx_passthru_kind to DEFAULT
gfx_passthru = "igd" => sets build_info.u.gfx_passthru to true
and build_info.u.gfx_passthru_kind to IGD

Here if gfx_passthru_kind = DEFAULT, we will call libxl__is_igd_vga_passthru() to check if we're hitting that table to need to pass that option to qemu. But if gfx_passthru_kind = "igd" we always force to pass that.

Quoted 5 lines

And "-gfx_passthru" is just introduced to work for qemu-xen-traditional so we should get this away from libxl__build_device_model_args_new() in the case of qemu upstream.

Acked + applied both patches, thanks.

Stefano -- are the QEMU side patches in qemu-upstream-unstable.git yet? If not I suppose this is a call/reminder to backport them from mainline or whatever.

Ian.

#7 lore raw

Re: [Qemu-devel] [v4][PATCH 2/2] libxl: introduce gfx_passthru_kind

Quoted 23 lines · Ian Campbell
Quoted 22 lines · Tiejun Chen

Although we already have 'gfx_passthru' in b_info, this doesn't suffice after we want to handle IGD specifically. Now we define a new field of type, gfx_passthru_kind, to indicate we're trying to pass IGD. Actually this means we can benefit this to support other specific devices just by extending gfx_passthru_kind. And then we can cooperate with gfx_passthru to address IGD cases as follows:

gfx_passthru = 0 => sets build_info.u.gfx_passthru to false
gfx_passthru = 1 => sets build_info.u.gfx_passthru to true and
build_info.u.gfx_passthru_kind to DEFAULT
gfx_passthru = "igd" => sets build_info.u.gfx_passthru to true
and build_info.u.gfx_passthru_kind to IGD

Here if gfx_passthru_kind = DEFAULT, we will call libxl__is_igd_vga_passthru() to check if we're hitting that table to need to pass that option to qemu. But if gfx_passthru_kind = "igd" we always force to pass that.

And "-gfx_passthru" is just introduced to work for qemu-xen-traditional so we should get this away from libxl__build_device_model_args_new() in the case of qemu upstream.

Quoted 7 lines
Quoted 1 line

Acked + applied both patches, thanks.

Stefano -- are the QEMU side patches in qemu-upstream-unstable.git yet? If not I suppose this is a call/reminder to backport them from mainline or whatever.

No, they are not, they'll be in qemu-upstream for Xen 4.7.

#8 lore raw
In reply to Tiejun Chen

Ping...

Thanks
Tiejun

Quoted 75 lines · Tiejun Chen

Ian,

As we discussed previously,

http://patchwork.ozlabs.org/patch/457055/

now it's time to push this into on xen/tools side since all qemu stuffs have been merged.

https://lists.gnu.org/archive/html/qemu-devel/2015-09/msg02094.html

v4:

Ian,

Actually we had v3.5 online previously, which was reviewed by you.

http://permalink.gmane.org/gmane.comp.emulators.qemu/329100

So here I just bring a little bit to refine code just for patch #2 according to out last conversation.

v3:

* Refine some codes based on Campbell's feedback so thanks for Campbell's kind guideline to patch #2
* Update the manpages in patch #2

v2:

* Refine patch #2's head description
* Improve codes quality inside patch #1 based on Wei's comments
* Refill the summary inside patch #0 based on Konrad and Wei's suggestion

When we're working to support IGD GFX passthrough with qemu upstream, instead of "-gfx_passthru" we'd like to make that a machine option, "-machine xxx,igd-passthru=on".

https://lists.nongnu.org/archive/html/qemu-devel/2015-01/msg02050.html

This need to bring a change on tool side.

After a discussion with Campbell, we'd like to construct a table to record all IGD devices we can support. If we hit that table, we should pass that option. And so we also introduce a new field of type, 'gfx_passthru_kind', to cooperate with 'gfx_passthru' to cover all scenarios like this,

gfx_passthru = 0 => sets build_info.u.gfx_passthru to false
gfx_passthru = 1 => sets build_info.u.gfx_passthru to true and
build_info.u.gfx_passthru_kind to DEFAULT
gfx_passthru = "igd" => sets build_info.u.gfx_passthru to false
and build_info.u.gfx_passthru_kind to IGD

And note actually that option "-gfx_passthru" is just introduced to work for qemu-xen-traditional so we should get this away from libxl__build_device_model_args_new() in the case of qemu upstream.

----------------------------------------------------------------
Tiejun Chen (2):
libxl: introduce libxl__is_igd_vga_passthru
libxl: introduce gfx_passthru_kind

  docs/man/xl.cfg.pod.5        |  35 ++++++++--
  tools/libxl/libxl.h          |   6 ++
  tools/libxl/libxl_dm.c       |  46 +++++++++++--
  tools/libxl/libxl_internal.h |   2 +
  tools/libxl/libxl_pci.c      | 124 +++++++++++++++++++++++++++++++++++
  tools/libxl/libxl_types.idl  |   6 ++
  tools/libxl/xl_cmdimpl.c     |  14 +++-
  7 files changed, 223 insertions(+), 10 deletions(-)

Thanks
Tiejun

#9 lore raw
In reply to Tiejun Chen

Ping...

Thanks
Tiejun

Quoted 75 lines · Tiejun Chen

Ian,

As we discussed previously,

http://patchwork.ozlabs.org/patch/457055/

now it's time to push this into on xen/tools side since all qemu stuffs have been merged.

https://lists.gnu.org/archive/html/qemu-devel/2015-09/msg02094.html

v4:

Ian,

Actually we had v3.5 online previously, which was reviewed by you.

http://permalink.gmane.org/gmane.comp.emulators.qemu/329100

So here I just bring a little bit to refine code just for patch #2 according to out last conversation.

v3:

* Refine some codes based on Campbell's feedback so thanks for Campbell's kind guideline to patch #2
* Update the manpages in patch #2

v2:

* Refine patch #2's head description
* Improve codes quality inside patch #1 based on Wei's comments
* Refill the summary inside patch #0 based on Konrad and Wei's suggestion

When we're working to support IGD GFX passthrough with qemu upstream, instead of "-gfx_passthru" we'd like to make that a machine option, "-machine xxx,igd-passthru=on".

https://lists.nongnu.org/archive/html/qemu-devel/2015-01/msg02050.html

This need to bring a change on tool side.

After a discussion with Campbell, we'd like to construct a table to record all IGD devices we can support. If we hit that table, we should pass that option. And so we also introduce a new field of type, 'gfx_passthru_kind', to cooperate with 'gfx_passthru' to cover all scenarios like this,

gfx_passthru = 0 => sets build_info.u.gfx_passthru to false
gfx_passthru = 1 => sets build_info.u.gfx_passthru to true and
build_info.u.gfx_passthru_kind to DEFAULT
gfx_passthru = "igd" => sets build_info.u.gfx_passthru to false
and build_info.u.gfx_passthru_kind to IGD

And note actually that option "-gfx_passthru" is just introduced to work for qemu-xen-traditional so we should get this away from libxl__build_device_model_args_new() in the case of qemu upstream.

----------------------------------------------------------------
Tiejun Chen (2):
libxl: introduce libxl__is_igd_vga_passthru
libxl: introduce gfx_passthru_kind

  docs/man/xl.cfg.pod.5        |  35 ++++++++--
  tools/libxl/libxl.h          |   6 ++
  tools/libxl/libxl_dm.c       |  46 +++++++++++--
  tools/libxl/libxl_internal.h |   2 +
  tools/libxl/libxl_pci.c      | 124 +++++++++++++++++++++++++++++++++++
  tools/libxl/libxl_types.idl  |   6 ++
  tools/libxl/xl_cmdimpl.c     |  14 +++-
  7 files changed, 223 insertions(+), 10 deletions(-)

Thanks
Tiejun