libxl: Allow multiple USB devices on HVM domain creation
A message that is not in the archive (sent off-list or not yet received): patchbomb.1354104964@elijah
PATCH 1/1 libxl: Allow multiple USB devices on HVM domain creation
# HG changeset patch
# User George Dunlap <[email protected]> # Date 1354101445 0
# Node ID 538d9ffbd71b41e8cf6d7da0ded9e0a0b07f3c0d # Parent ae6fb202b233af815466055d9f1a635802a50855
libxl: Allow multiple USB devices on HVM domain creation
This patch allows an HVM domain to be created with multiple USB devices.
Since the previous interface only allowed the passing of a single device, this requires us to add a new element to the hvm struct of libxl_domain_build_info -- usbdevice_list. For API compatibility, the old element, usbdevice, remains.
If b_info->hvm.usbdevice is non-NULL, then it will be used exclusively; otherwise, libxl will check for b_info->hvm.usbdevice_list instead. Each device listed will cause an extra "-usbdevice [foo]" to be appended to the qemu command line.
In order to allow users of libxl to write software compatible with older versions of libxl, also define LIBXL_HAVE_BUILDINFO_USBDEVICE_LIST. If this is present, the caller should use b_info->hvm.usbdevice_list; otherwise they should use b_info->hvm.usbdevice.
tools/libxl/libxl.h +17 -0
@@ -266,6 +266,23 @@#endif #endif +/* +* LIBXL_HAVE_BUILDINFO_USBDEVICE_LIST +* +* If this is defined, then the libxl_domain_build_info structure will +* contain hvm.usbdevice_list, a libxl_string_list type that contains +* a list of USB devices to specify on the qemu command-line. +* +* If it is set, callers may use either hvm.usbdevice or +* hvm.usbdevice_list, but not both; if both are set, only +* hvm.usbdevice will be used. +* +* If this is not defined, callers should only use hvm.usbdevice. +* Note that this means only one device can be added at domain build +* time. +*/ +#define LIBXL_HAVE_BUILDINFO_USBDEVICE_LIST 1 +/* Functions annotated with LIBXL_EXTERNAL_CALLERS_ONLY may not be * called from within libxl itself. Callers outside libxl, who * do not #include libxl_internal.h, are fine. */
tools/libxl/libxl_dm.c +9 -0
@@ -191,6 +191,15 @@ static char ** libxl__build_device_modelif (b_info->u.hvm.usbdevice) {flexarray_vappend (dm_args,"-usbdevice" , b_info->u.hvm.usbdevice,NULL ); + }else if (b_info->u.hvm.usbdevice_list) { +char **p; +for (p = b_info->u.hvm.usbdevice_list; + *p; + p++) { +flexarray_vappend (dm_args, +"-usbdevice" , + *p,NULL ); + } } }if (b_info->u.hvm.soundhw) {
tools/libxl/libxl_types.idl +1 -0
@@ -326,6 +326,7 @@ libxl_domain_build_info = Struct("domain ("usbdevice", string), ("soundhw", string), ("xen_platform_pci", libxl_defbool), + ("usbdevice_list", libxl_string_list), ])), ("pv", Struct(None, [("kernel", string), ("slack_memkb", MemKB),
Quoted 10 lines · George Dunlap
# HG changeset patch
# User George Dunlap <[email protected]> # Date 1354101445 0
# Node ID538d9ffbd71b41e8cf6d7da0ded9e0a0b07f3c0d# Parentae6fb202b233af815466055d9f1a635802a50855
libxl: Allow multiple USB devices on HVM domain creationThis patch allows an HVM domain to be created with multiple USB devices.
Quoted 7 lines
Since the previous interface only allowed the passing of a single device, this requires us to add a new element to the hvm struct of
libxl_domain_build_info--usbdevice_list. For API compatibility, the old element, usbdevice, remains.If
b_info->hvm.usbdeviceis non-NULL, then it will be used exclusively; otherwise, libxl will check forb_info->hvm.usbdevice_listinstead.
Is this the right way round? If the caller knows about usbdevice_list enough to have set it to something then surely that's the one it wanted?
Quoted 7 lines
Each device listed will cause an extra "-usbdevice [foo]" to be appended to the qemu command line.
In order to allow users of libxl to write software compatible with older versions of libxl, also define
LIBXL_HAVE_BUILDINFO_USBDEVICE_LIST. If this is present, the caller should useb_info->hvm.usbdevice_list; otherwise they should useb_info->hvm.usbdevice.
Actually it's a both stricter and looser than that, if LIBXL_HAVE_BUILDINFO_USBDEVICE_LIST is defined then the caller can choose to use usbdevice_list but can also use usbdevice if they wish (but not both). If it is not present then they must not use usbdevice_list at all.
I wonder if this LIBXL_HAVE should also be contained in a #ifdef LIBXL_API_VERSION >= 0x040300?
Quoted 66 lines
Signed-off-by George Dunlap <[email protected]>tools/libxl/libxl.h +17
-0@@ -266,6 +266,23 @@#endif #endif +/* +* LIBXL_HAVE_BUILDINFO_USBDEVICE_LIST +* +* If this is defined, then the libxl_domain_build_info structure will +* contain hvm.usbdevice_list, a libxl_string_list type that contains +* a list of USB devices to specify on the qemu command-line. +* +* If it is set, callers may use either hvm.usbdevice or +* hvm.usbdevice_list, but not both; if both are set, only +* hvm.usbdevice will be used. +* +* If this is not defined, callers should only use hvm.usbdevice. +* Note that this means only one device can be added at domain build +* time. +*/ +#define LIBXL_HAVE_BUILDINFO_USBDEVICE_LIST 1 +/* Functions annotated with LIBXL_EXTERNAL_CALLERS_ONLY may not be * called from within libxl itself. Callers outside libxl, who * do not #include libxl_internal.h, are fine. */ tools/libxl/libxl_dm.c +9
-0@@ -191,6 +191,15 @@ static char ** libxl__build_device_modelif (b_info->u.hvm.usbdevice) {flexarray_vappend (dm_args,"-usbdevice" , b_info->u.hvm.usbdevice,NULL ); + }else if (b_info->u.hvm.usbdevice_list) { +char **p; +for (p = b_info->u.hvm.usbdevice_list; + *p; + p++) { +flexarray_vappend (dm_args, +"-usbdevice" , + *p,NULL ); + } } }if (b_info->u.hvm.soundhw) {tools/libxl/libxl_types.idl +1
-0@@ -326,6 +326,7 @@ libxl_domain_build_info = Struct("domain ("usbdevice", string), ("soundhw", string), ("xen_platform_pci", libxl_defbool), + ("usbdevice_list", libxl_string_list), ])), ("pv", Struct(None, [("kernel", string), ("slack_memkb", MemKB),_______________________________________________
Xen-devel mailing list
[email protected]
http://lists.xen.org/xen-devel
Quoted 16 lines · Ian Campbell
Quoted 15 lines · George Dunlap
# HG changeset patch
# User George Dunlap <[email protected]> # Date 1354101445 0
# Node ID538d9ffbd71b41e8cf6d7da0ded9e0a0b07f3c0d# Parentae6fb202b233af815466055d9f1a635802a50855
libxl: Allow multiple USB devices on HVM domain creationThis patch allows an HVM domain to be created with multiple USB devices.
Since the previous interface only allowed the passing of a single device, this requires us to add a new element to the hvm struct of
libxl_domain_build_info--usbdevice_list. For API compatibility, the old element, usbdevice, remains.
Quoted 4 lines
Quoted 2 lines
If
b_info->hvm.usbdeviceis non-NULL, then it will be used exclusively; otherwise, libxl will check forb_info->hvm.usbdevice_listinstead.Is this the right way round? If the caller knows about
usbdevice_listenough to have set it to something then surely that's the one it wanted?
Well, I had considered returning an error if both were set. :-) I
suppose actually that's probably the way that is likely to flag up bugs in the toolstack the best -- only doing one or the other will cause weird buggy behavior that will be hard to track down. Does that sound good to you?
Quoted 7 lines
Quoted 7 lines
Each device listed will cause an extra "-usbdevice [foo]" to be appended to the qemu command line.
In order to allow users of libxl to write software compatible with older versions of libxl, also define
LIBXL_HAVE_BUILDINFO_USBDEVICE_LIST. If this is present, the caller should useb_info->hvm.usbdevice_list; otherwise they should useb_info->hvm.usbdevice.
Quoted 5 lines
Actually it's a both stricter and looser than that, if
LIBXL_HAVE_BUILDINFO_USBDEVICE_LISTis defined then the caller can choose to useusbdevice_listbut can also use usbdevice if they wish (but not both). If it is not present then they must not useusbdevice_listat all.
Hmm, I think I wrote this first and then changed it and forgot to update it. :-) I also forgot that in this kind of "API spec" document, "should" and "must" are technical terms with very specific definitions. I'll revise it when I re-spin.
Quoted 3 lines
I wonder if this
LIBXL_HAVEshould also be contained in a #ifdefLIBXL_API_VERSION>=0x040300?
What would be the purpose of that?
-George
Quoted 41 lines · George Dunlap
Quoted 16 lines · Ian Campbell
Quoted 15 lines · George Dunlap
# HG changeset patch
# User George Dunlap <[email protected]> # Date 1354101445 0
# Node ID538d9ffbd71b41e8cf6d7da0ded9e0a0b07f3c0d# Parentae6fb202b233af815466055d9f1a635802a50855
libxl: Allow multiple USB devices on HVM domain creationThis patch allows an HVM domain to be created with multiple USB devices.
Since the previous interface only allowed the passing of a single device, this requires us to add a new element to the hvm struct of
libxl_domain_build_info--usbdevice_list. For API compatibility, the old element, usbdevice, remains.Quoted 4 lines
Quoted 2 lines
If
b_info->hvm.usbdeviceis non-NULL, then it will be used exclusively; otherwise, libxl will check forb_info->hvm.usbdevice_listinstead.Is this the right way round? If the caller knows about
usbdevice_listenough to have set it to something then surely that's the one it wanted?Well, I had considered returning an error if both were set. :-) I
suppose actually that's probably the way that is likely to flag up bugs in the toolstack the best -- only doing one or the other will cause weird buggy behavior that will be hard to track down. Does that sound good to you?Quoted 7 lines
Quoted 7 lines
Each device listed will cause an extra "-usbdevice [foo]" to be appended to the qemu command line.
In order to allow users of libxl to write software compatible with older versions of libxl, also define
LIBXL_HAVE_BUILDINFO_USBDEVICE_LIST. If this is present, the caller should useb_info->hvm.usbdevice_list; otherwise they should useb_info->hvm.usbdevice.Quoted 5 lines
Actually it's a both stricter and looser than that, if
LIBXL_HAVE_BUILDINFO_USBDEVICE_LISTis defined then the caller can choose to useusbdevice_listbut can also use usbdevice if they wish (but not both). If it is not present then they must not useusbdevice_listat all.
Quoted 4 lines
Hmm, I think I wrote this first and then changed it and forgot to update it. :-) I also forgot that in this kind of "API spec" document, "should" and "must" are technical terms with very specific definitions. I'll revise it when I re-spin.
I don't think we actually use the RFC-whichever definitions of should and must in general (and they would be SHOULD and MUST anyhow), it's probably me being picky.
Quoted 4 lines
Quoted 2 lines
I wonder if this
LIBXL_HAVEshould also be contained in a #ifdefLIBXL_API_VERSION>=0x040300?What would be the purpose of that?
LIBXL_API_VERSION and LIBXL_HAVE_* are two different mechanisms for helping us to do API compatibility, I hadn't properly thought about the interaction between them until now.
Adding the ifdef would expose the new feature only to those claiming compatibility with the newer API.
More generally the purpose if LIBXL_API_VERSION generally is to allow us to deprecate interfaces over time, which isn't quite the case here I suppose.
Ian.
PATCH 1/1 xl: Accept a list for usbdevice in config file
In reply to a message not in the archive# HG changeset patch
# User George Dunlap <[email protected]> # Date 1354104851 0
# Node ID e87d8fad28e00090591b65b2682dea9112f8830d # Parent 538d9ffbd71b41e8cf6d7da0ded9e0a0b07f3c0d
xl: Accept a list for usbdevice in config file
Allow the "usbdevice" key to accept a list of USB devices, and pass them in using the new usbdevice_list domain build element.
For backwards compatibility, still accept singleton values.
Also update the xl.cfg manpage, adding information about how to pass through host devices.
docs/man/xl.cfg.pod.5 +15 -8
@@ -1103,15 +1103,22 @@ device. Enables or disables a USB bus in the guest. -=item B<usbdevice=DEVICE> +=item B<usbdevice=[ "DEVICE", "DEVICE", ...]> -Adds B<DEVICE> to the USB bus. The USB bus must also be enabled using -B<usb=1>. The most common use for this option is B<usbdevice=tablet> -which adds pointer device using absolute coordinates. Such devices -function better than relative coordinate devices (such as a standard -mouse) since many methods of exporting guest graphics (such as VNC) -work better in this mode. Note that this is independent of the actual -pointer device you are using on the host/client side. +Adds B<DEVICE>s to the emulated USB bus. The USB bus must also be +enabled using B<usb=1>. The most common use for this option is +B<usbdevice=['tablet']> which adds pointer device using absolute +coordinates. Such devices function better than relative coordinate +devices (such as a standard mouse) since many methods of exporting +guest graphics (such as VNC) work better in this mode. Note that this +is independent of the actual pointer device you are using on the +host/client side. + +Host devices can also be passed through in this way, by specifying +host:USBID, where USBID is of the form xxxx:yyyy. The USBID can +typically be found by using lsusb or usb-devices. + +The form usbdevice=DEVICE is also accepted for backwards compatibility. =back
tools/libxl/xl_cmdimpl.c +17 -2
@@ -1485,8 +1485,23 @@ skip_vfb:xlu_cfg_replace_string (config,"serial" , &b_info->u.hvm.serial,0 );xlu_cfg_replace_string (config,"boot" , &b_info->u.hvm.boot,0 );xlu_cfg_get_defbool (config,"usb" , &b_info->u.hvm.usb,0 ); -xlu_cfg_replace_string (config,"usbdevice" , - &b_info->u.hvm.usbdevice,0 ); +switch (xlu_cfg_get_list_as_string_list (config,"usbdevice" , + &b_info->u.hvm.usbdevice_list, +1 )) + { + +case 0 :break ;/* Success */ +case ESRCH :break ;/* Option not present */ +case EINVAL : +/* If it's not a valid list, try reading it as an atom, falling through to +* an error if it fails */ +if (!xlu_cfg_replace_string (config,"usbdevice" , &b_info->u.hvm.usbdevice,0 )) +break ; +/* FALLTHRU */ +default : +fprintf (stderr,"xl: Unable to parse usbdevice.\n" ); +exit (-ERROR_FAIL ); + }xlu_cfg_replace_string (config,"soundhw" , &b_info->u.hvm.soundhw,0 );xlu_cfg_get_defbool (config,"xen_platform_pci" , &b_info->u.hvm.xen_platform_pci,0 );
Re: [PATCH 2 of 2] xl: Accept a list for usbdevice in config file
Quoted 21 lines · George Dunlap
# HG changeset patch
# User George Dunlap <[email protected]> # Date 1354104851 0
# Node IDe87d8fad28e00090591b65b2682dea9112f8830d# Parent538d9ffbd71b41e8cf6d7da0ded9e0a0b07f3c0d
xl: Accept a list for usbdevice in config fileAllow the "usbdevice" key to accept a list of USB devices, and pass them in using the new
usbdevice_listdomain build element.For backwards compatibility, still accept singleton values.
Also update the xl.cfg manpage, adding information about how to pass through host devices.
Signed-off-by George Dunlap <[email protected]>
docs/man/xl.cfg.pod.5 device. · line 1105
Quoted 3 lines
docs/man/xl.cfg.pod.5 +0
-0@@ -1103,15 +1103,22 @@ device. Enables or disables a USB bus in the guest.
Should we add "HVM" there ?
Quoted 15 lines
docs/man/xl.cfg.pod.5 +6
-8-=item B<usbdevice=DEVICE> +=item B<usbdevice=[ "DEVICE", "DEVICE", ...]> -Adds B<DEVICE> to the USB bus. The USB bus must also be enabled using -B<usb=1>. The most common use for this option is B<usbdevice=tablet> -which adds pointer device using absolute coordinates. Such devices -function better than relative coordinate devices (such as a standard -mouse) since many methods of exporting guest graphics (such as VNC) -work better in this mode. Note that this is independent of the actual -pointer device you are using on the host/client side. +Adds B<DEVICE>s to the emulated USB bus. The USB bus must also be +enabled using B<usb=1>. The most common use for this option is +B<usbdevice=['tablet']> which adds pointer device using absolute +coordinates. Such devices function better than relative coordinate +devices (such as a standard mouse) since many methods of exporting
docs/man/xl.cfg.pod.5 device. · line 1122
Quoted 10 lines
docs/man/xl.cfg.pod.5 +9
-0+guest graphics (such as VNC) work better in this mode. Note that this +is independent of the actual pointer device you are using on the +host/client side. + +Host devices can also be passed through in this way, by specifying +host:USBID, where USBID is of the form xxxx:yyyy. The USBID can +typically be found by using lsusb or usb-devices. + +The form usbdevice=DEVICE is also accepted for backwards compatibility.
Should we mention the difference between qemu-traditional and qemu-upstream here?
qemu-traditional only supports USB 1.1 devices, while qemu-upstream offers USB 2.0 (and 3.0) emulation. Is there currently a way to control to which virtual usb controller the host device gets added to?
See here for help how to configure that with qemu/kvm:
http://www.linux-kvm.com/content/qemu-kvm-11-adds-experimental-support-usb-30
-- Pasi
Re: [PATCH 2 of 2] xl: Accept a list for usbdevice in config file
Quoted 23 lines · Pasi Kärkkäinen
Quoted 22 lines · George Dunlap
# HG changeset patch
# User George Dunlap <[email protected]> # Date 1354104851 0
# Node IDe87d8fad28e00090591b65b2682dea9112f8830d# Parent538d9ffbd71b41e8cf6d7da0ded9e0a0b07f3c0d
xl: Accept a list for usbdevice in config fileAllow the "usbdevice" key to accept a list of USB devices, and pass them in using the new
usbdevice_listdomain build element.For backwards compatibility, still accept singleton values.
Also update the xl.cfg manpage, adding information about how to pass through host devices.
Signed-off-by George Dunlap <[email protected]>docs/man/xl.cfg.pod.5 +0
-0@@ -1103,15 +1103,22 @@ device.
Quoted 5 lines
docs/man/xl.cfg.pod.5 device. · line 1105
Quoted 2 lines
docs/man/xl.cfg.pod.5 +0
-0Enables or disables a USB bus in the guest.Should we add "HVM" there ?
I could make this say, "...in an HVM guest". (I'll wait to see if I get other comments before re-sending the patch.)
Quoted 31 lines
Quoted 17 lines
docs/man/xl.cfg.pod.5 +6
-8-=item B<usbdevice=DEVICE> +=item B<usbdevice=[ "DEVICE", "DEVICE", ...]> -Adds B<DEVICE> to the USB bus. The USB bus must also be enabled using -B<usb=1>. The most common use for this option is B<usbdevice=tablet> -which adds pointer device using absolute coordinates. Such devices -function better than relative coordinate devices (such as a standard -mouse) since many methods of exporting guest graphics (such as VNC) -work better in this mode. Note that this is independent of the actual -pointer device you are using on the host/client side. +Adds B<DEVICE>s to the emulated USB bus. The USB bus must also be +enabled using B<usb=1>. The most common use for this option is +B<usbdevice=['tablet']> which adds pointer device using absolute +coordinates. Such devices function better than relative coordinate +devices (such as a standard mouse) since many methods of exportingdocs/man/xl.cfg.pod.5 device. · line 1122
Quoted 10 lines
docs/man/xl.cfg.pod.5 +9
-0+guest graphics (such as VNC) work better in this mode. Note that this +is independent of the actual pointer device you are using on the +host/client side. + +Host devices can also be passed through in this way, by specifying +host:USBID, where USBID is of the form xxxx:yyyy. The USBID can +typically be found by using lsusb or usb-devices. + +The form usbdevice=DEVICE is also accepted for backwards compatibility.Should we mention the difference between qemu-traditional and qemu-upstream here?
Quoted 9 lines
qemu-traditional only supports USB 1.1 devices, while qemu-upstream offers USB 2.0 (and 3.0) emulation.
Is there currently a way to control to which virtual usb controller the host device gets added to?See here for help how to configure that with qemu/kvm:
http://www.linux-kvm.com/content/qemu-kvm-11-adds-experimental-support-usb-30
I think this level of detail is probably too high for a man page -- this kind of thing should be on a wiki somewhere I think.
In general though, I think xl's approach to USB stuff is really simplistic. It would be nice to have it be more fully-featured, but I'm not sure that's on anyone's priority list.
-George
_______________________________________________
Xen-devel mailing list
[email protected]
http://lists.xen.org/xen-devel
Re: [PATCH 2 of 2] xl: Accept a list for usbdevice in config file
Quoted 32 lines · George Dunlap
On Wed, Nov 28, 2012 at 3:11 PM, Pasi Kärkkäinen <[email protected]> wrote:
On Wed, Nov 28, 2012 at 12:16:06PM +0000, George Dunlap wrote: > # HG changeset patch > # User George Dunlap <[email protected]> > # Date 1354104851 0 > # Node ID e87d8fad28e00090591b65b2682dea9112f8830d > # Parent 538d9ffbd71b41e8cf6d7da0ded9e0a0b07f3c0d > xl: Accept a list for usbdevice in config file > > Allow the "usbdevice" key to accept a list of USB devices, and pass > them in using the new usbdevice_list domain build element. > > For backwards compatibility, still accept singleton values. > > Also update the xl.cfg manpage, adding information about how to pass > through host devices. > > Signed-off-by: George Dunlap <[email protected]> > > diff --git a/docs/man/xl.cfg.pod.5 b/docs/man/xl.cfg.pod.5 > --- a/docs/man/xl.cfg.pod.5 > +++ b/docs/man/xl.cfg.pod.5 > @@ -1103,15 +1103,22 @@ device. > > Enables or disables a USB bus in the guest. > Should we add "HVM" there ?
Quoted 2 lines
I could make this say, "...in an HVM guest". (I'll wait to see if I get other comments before re-sending the patch.)
FWIW other similar options simply begin the paragraph with "(HVM only)"
But actually this particular option is already underneath a section which begins "The following options apply only to HVM guests." and we don't seem to reiterate that in the other options.
Quoted 7 lines
> -=item B<usbdevice=DEVICE>
> +=item B<usbdevice=[ "DEVICE", "DEVICE", ...]> >
> -Adds B<DEVICE> to the USB bus. The USB bus must also be enabled using
> -B<usb=1>. The most common use for this option is B<usbdevice=tablet>
Are there other non-host* options?
Quoted 2 lines
I think this level of detail is probably too high for a man page -- this kind of thing should be on a wiki somewhere I think.
Its a bit of a cop-out but we do reference qemu(1) in some other places, which might suffice here?
Ian.
_______________________________________________
Xen-devel mailing list
[email protected]
http://lists.xen.org/xen-devel
Re: [PATCH 2 of 2] xl: Accept a list for usbdevice in config file
Quoted 33 lines · Ian Campbell
Quoted 32 lines · George Dunlap
On Wed, Nov 28, 2012 at 3:11 PM, Pasi Kärkkäinen <[email protected]> wrote:
On Wed, Nov 28, 2012 at 12:16:06PM +0000, George Dunlap wrote: > # HG changeset patch > # User George Dunlap <[email protected]> > # Date 1354104851 0 > # Node ID e87d8fad28e00090591b65b2682dea9112f8830d > # Parent 538d9ffbd71b41e8cf6d7da0ded9e0a0b07f3c0d > xl: Accept a list for usbdevice in config file > > Allow the "usbdevice" key to accept a list of USB devices, and pass > them in using the new usbdevice_list domain build element. > > For backwards compatibility, still accept singleton values. > > Also update the xl.cfg manpage, adding information about how to pass > through host devices. > > Signed-off-by: George Dunlap <[email protected]> > > diff --git a/docs/man/xl.cfg.pod.5 b/docs/man/xl.cfg.pod.5 > --- a/docs/man/xl.cfg.pod.5 > +++ b/docs/man/xl.cfg.pod.5 > @@ -1103,15 +1103,22 @@ device. > > Enables or disables a USB bus in the guest. > Should we add "HVM" there ?
Quoted 7 lines
Quoted 2 lines
I could make this say, "...in an HVM guest". (I'll wait to see if I get other comments before re-sending the patch.)
FWIW other similar options simply begin the paragraph with "(HVM only)"
But actually this particular option is already underneath a section which begins "The following options apply only to HVM guests." and we don't seem to reiterate that in the other options.
Ah, right -- now I remember, I thought about it, saw that it was in an "HVM only" section, and decided that was sufficient. :-)
Quoted 9 lines
Quoted 8 lines
> -=item B<usbdevice=DEVICE>
> +=item B<usbdevice=[ "DEVICE", "DEVICE", ...]> >
> -Adds B<DEVICE> to the USB bus. The USB bus must also be enabled using
> -B<usb=1>. The most common use for this option is B<usbdevice=tablet>Are there other non-host* options?
qemu.git/vl.c suggests other options might include, "disk:", "serial:", "net:" and "bt:"; other non-colon-including options include "keyboard", "mouse", "wacom-tablet", and "braile".
On the other hand, a number of these seem to require supplementary options to work well; e.g., network and bt (bluetooth) require one argument to hook up a USB device, and another option to say how it's connected to what.
It's also interesting to note that the KVM documentation doesnt' mention using "-usb" and "-usbdevice" at all -- they seem to prefer using "-device" to specify hubs &c.
But still, it might be useful to say that these are passed via "usbdevice" to the qemu command line, and so you can find more information on the qemu man page.
-George
_______________________________________________
Xen-devel mailing list
[email protected]
http://lists.xen.org/xen-devel
Re: [PATCH 2 of 2] xl: Accept a list for usbdevice in config file
On Thu, Nov 29, 2012 at 11:44 AM, George Dunlap <[email protected]
Quoted 6 lines
wrote:
qemu.git/vl.csuggests other options might include, "disk:", "serial:", "net:" and "bt:"; other non-colon-including options include "keyboard", "mouse", "wacom-tablet", and "braile".
Quoted 8 lines
On the other hand, a number of these seem to require supplementary options to work well; e.g., network and bt (bluetooth) require one argument to hook up a USB device, and another option to say how it's connected to what.
It's also interesting to note that the KVM documentation doesnt' mention using "-usb" and "-usbdevice" at all -- they seem to prefer using "-device" to specify hubs &c.
Speaking of which, the KVM docs refer to "-usb" and "-usbdevice" as "legacy interfaces", which will only get you piix3. It seems like exposing the full capabilities of qemu would mean either 1) coming up with a full specifciation which we can then translate into qemu directives, as libvirt seems to do, or 2) just recommend people construct their own qemu command-line options to pass through.
It's not 100% clear to me that 1 is actually less complicated from a user perspective in the long run than 2, and it's certainly a lot more work...
-George
_______________________________________________
Xen-devel mailing list
[email protected]
http://lists.xen.org/xen-devel
Re: [PATCH 2 of 2] xl: Accept a list for usbdevice in config file
Quoted 11 lines · George Dunlap
On Thu, Nov 29, 2012 at 11:44 AM, George Dunlap <[email protected]> wrote:
qemu.git/vl.csuggests other options might include, "disk:", "serial:", "net:" and "bt:"; other non-colon-including options include "keyboard", "mouse", "wacom-tablet", and "braile".On the other hand, a number of these seem to require supplementary options to work well; e.g., network and bt (bluetooth) require one argument to hook up a USB device, and another option to say how it's connected to what.
Quoted 6 lines
It's also interesting to note that the KVM documentation doesnt' mention using "-usb" and "-usbdevice" at all -- they seem to prefer using "-device" to specify hubs &c.
Speaking of which, the KVM docs refer to "-usb" and "-usbdevice" as "legacy interfaces", which will only get you piix3.
This is a concern for other types of device too. We seem to invoke new qemu with a mixture of legacy and new-style interfaces.
Quoted 5 lines
It seems like exposing the full capabilities of qemu would mean
either 1) coming up with a full specifciation which we can then translate into qemu directives, as libvirt seems to do, or 2) just recommend people construct their own qemu command-line options to pass through.
Does libvirt let you care about topologies or just it just automatically create a new USB controller for every N devices you add and hook things up in some order?
Quoted 6 lines
It's not 100% clear to me that 1 is actually less complicated from a user perspective in the long run than 2, and it's certainly a lot more work...
-George
Re: [PATCH 2 of 2] xl: Accept a list for usbdevice in config file
Quoted 21 lines · Ian Campbell
Quoted 11 lines · George Dunlap
On Thu, Nov 29, 2012 at 11:44 AM, George Dunlap <[email protected]> wrote:
qemu.git/vl.csuggests other options might include, "disk:", "serial:", "net:" and "bt:"; other non-colon-including options include "keyboard", "mouse", "wacom-tablet", and "braile".On the other hand, a number of these seem to require supplementary options to work well; e.g., network and bt (bluetooth) require one argument to hook up a USB device, and another option to say how it's connected to what.
Quoted 6 lines
It's also interesting to note that the KVM documentation doesnt' mention using "-usb" and "-usbdevice" at all -- they seem to prefer using "-device" to specify hubs &c.
Speaking of which, the KVM docs refer to "-usb" and "-usbdevice" as "legacy interfaces", which will only get you piix3.
This is a concern for other types of device too. We seem to invoke new qemu with a mixture of legacy and new-style interfaces.
Quoted 8 lines
Quoted 5 lines
It seems like exposing the full capabilities of qemu would mean
either 1) coming up with a full specifciation which we can then translate into qemu directives, as libvirt seems to do, or 2) just recommend people construct their own qemu command-line options to pass through.Does libvirt let you care about topologies or just it just automatically create a new USB controller for every N devices you add and hook things up in some order?
No idea -- I took a quick glance through the libvirt docs and I didn't see anything particular about a topology. I don't have a system set up with libvirt, so I can't try it and see what the resulting qemu command line looks like :-)
http://libvirt.org/formatdomain.html#elementsUSB
-George
Re: [PATCH 2 of 2] xl: Accept a list for usbdevice in config file
In reply to Ian CampbellQuoted 28 lines · Ian Campbell
Quoted 11 lines · George Dunlap
On Thu, Nov 29, 2012 at 11:44 AM, George Dunlap <[email protected]> wrote:
qemu.git/vl.csuggests other options might include, "disk:", "serial:", "net:" and "bt:"; other non-colon-including options include "keyboard", "mouse", "wacom-tablet", and "braile".On the other hand, a number of these seem to require supplementary options to work well; e.g., network and bt (bluetooth) require one argument to hook up a USB device, and another option to say how it's connected to what.
Quoted 6 lines
It's also interesting to note that the KVM documentation doesnt' mention using "-usb" and "-usbdevice" at all -- they seem to prefer using "-device" to specify hubs &c.
Speaking of which, the KVM docs refer to "-usb" and "-usbdevice" as "legacy interfaces", which will only get you piix3.
This is a concern for other types of device too. We seem to invoke new qemu with a mixture of legacy and new-style interfaces.
Quoted 5 lines
It seems like exposing the full capabilities of qemu would mean
either 1) coming up with a full specifciation which we can then translate into qemu directives, as libvirt seems to do, or 2) just recommend people construct their own qemu command-line options to pass through.
Quoted 4 lines
Does libvirt let you care about topologies or just it just automatically create a new USB controller for every N devices you add and hook things up in some order?
Not sure about libvirt.. quick googling suggests that libvirt doesn't allow specifying the usb controller.
with qemu-kvm you can assign host usb devices to specific usb controllers using the "bus=" option:
http://www.linux-kvm.com/content/qemu-kvm-11-adds-experimental-support-usb-30
qemu-kvm -device nec-usb-xhci,id=xhci -device usb-storage,bus=xhci.0,drive=usbstick
-- Pasi