Re: [Xen-devel] [PATCH v2] libxl: Add AHCI support for upstream qemu

[Ian Campbell <ian.campbell@xxxxxxxxxx>]

On Tue, 2015-06-23 at 11:15 +0200, Fabio Fantoni wrote:
> Usage:
> ahci=0|1 (default=0)

I think a global rather than per disk option is OK (I can't think why a
user would want to mix and match) but maybe we should consider using an
enum (with values ide and ahci, defaulting to ide in libxl) so that we
can add support for whatever fancy new disk controller everyone is using
in 5 years time?

> If enabled adds ich9 disk controller in ahci mode and uses it with
> upstream qemu to emulate disks instead of ide.
> It doesn't support cdroms which still using ide (cdroms will use
> "-device ide-cd" as new qemu parameter)

I don't follow this reference to "will use" and a new qemu parameter,
there seems to be nothing corresponding in this patch AFAICT.

> Ahci requires new qemu parameter but for now other emulated disks cases
> remains with old ones (I did it in other patch, not needed by this one)

You can drop the reference to the other patch I think.

> I did it as libxl parameter disabled by default to avoid possible
> problems:
> - with save/restore/migration (restoring with ahci a domU that was with
> ide instead)
> - windows < 8 without pv drivers (a registry key change is needed for
> AHCI<->IDE change FWIK to avoid possible blue screen)

What is "FWIK"?

> - windows XP or older that many not support ahci by default.
> Setting AHCI with libxl parameter and default to disabled seems the best
> solution.
> AHCI increase hvm domUs boot performance. On linux hvm domU I saw up to
> only 20% of the previous total boot time, whereas boot time decrease a
> lot on W7 domUs for most of boots I have done. Small difference in boot
> time compared to ide mode on W8 and newer (probably other xen
> improvements or fixes are needed not ahci related)
> 
> Signed-off-by: Fabio Fantoni <fabio.fantoni@xxxxxxx>
> 
> ---
> 
> Changes in v2:
> - libxl_dm.c: small code style fix
> - added vbd-interface.txt changes
> ---
>  docs/man/xl.cfg.pod.5       |  9 +++++++++
>  docs/misc/vbd-interface.txt |  5 +++--
>  tools/libxl/libxl.h         | 10 ++++++++++
>  tools/libxl/libxl_create.c  |  1 +
>  tools/libxl/libxl_dm.c      | 10 +++++++++-
>  tools/libxl/libxl_types.idl |  1 +
>  tools/libxl/xl_cmdimpl.c    |  1 +
>  7 files changed, 34 insertions(+), 3 deletions(-)
> 
> diff --git a/docs/man/xl.cfg.pod.5 b/docs/man/xl.cfg.pod.5
> index a3e0e2e..7e16123 100644
> --- a/docs/man/xl.cfg.pod.5
> +++ b/docs/man/xl.cfg.pod.5
> @@ -904,6 +904,15 @@ default is B<cd>.
>  
>  =back
>  
> +=item B<ahci=[0|1]>

"=item B<ahci=BOOLEAN>" please.

> +If enabled adds ich9 disk controller in ahci mode and uses it with
> +upstream qemu to emulate disks instead of ide. It decrease boot time but

"decreases"

> +may be not supported by default in windows xp and older windows.
> +The default is disabled (0).

"may not be supported".

I think AHCI and IDE should be capitalised in the text (not the option
name). As should "Windows XP" and "Windows"

> +
> +=back
> +
>  =head3 Paging
>  
>  The following options control the mechanisms used to virtualise guest
> diff --git a/docs/misc/vbd-interface.txt b/docs/misc/vbd-interface.txt
> index f873db0..afb6846 100644
> --- a/docs/misc/vbd-interface.txt
> +++ b/docs/misc/vbd-interface.txt
> @@ -3,18 +3,19 @@ Xen guest interface
>  
>  A Xen guest can be provided with block devices.  These are always
>  provided as Xen VBDs; for HVM guests they may also be provided as
> -emulated IDE or SCSI disks.
> +emulated IDE, AHCI or SCSI disks.
>  
>  The abstract interface involves specifying, for each block device:
>  
>   * Nominal disk type: Xen virtual disk (aka xvd*, the default); SCSI
> -   (sd*); IDE (hd*).
> +   (sd*); IDE or AHCI (hd*).
>  
>     For HVM guests, each whole-disk hd* and and sd* device is made
>     available _both_ via emulated IDE resp. SCSI controller, _and_ as a
>     Xen VBD.  The HVM guest is entitled to assume that the IDE or SCSI
>     disks available via the emulated IDE controller target the same
>     underlying devices as the corresponding Xen VBD (ie, multipath).
> +   In hd* case with ahci=1, disk will be AHCI via emulated ich9 controller.
>  
>     For PV guests every device is made available to the guest only as a
>     Xen VBD.  For these domains the type is advisory, for use by the
> diff --git a/tools/libxl/libxl.h b/tools/libxl/libxl.h
> index 0a7913b..6a3677d 100644
> --- a/tools/libxl/libxl.h
> +++ b/tools/libxl/libxl.h
> @@ -596,6 +596,16 @@ typedef struct libxl__ctx libxl_ctx;
>  #define LIBXL_HAVE_SPICE_STREAMINGVIDEO 1
>  
>  /*
> + * LIBXL_HAVE_AHCI
> + *
> + * If defined, then the u.hvm structure will contain a boolean type:
> + * ahci. This value defines if ahci support is present.
> + *
> + * If this is not defined, the ahci support is ignored.
> + */
> +#define LIBXL_HAVE_AHCI 1
> +
> +/*
>   * LIBXL_HAVE_DOMAIN_CREATE_RESTORE_PARAMS 1
>   *
>   * If this is defined, libxl_domain_create_restore()'s API has changed to
> diff --git a/tools/libxl/libxl_create.c b/tools/libxl/libxl_create.c
> index 86384d2..8ca2481 100644
> --- a/tools/libxl/libxl_create.c
> +++ b/tools/libxl/libxl_create.c
> @@ -331,6 +331,7 @@ int libxl__domain_build_info_setdefault(libxl__gc *gc,
>          libxl_defbool_setdefault(&b_info->u.hvm.nested_hvm,         false);
>          libxl_defbool_setdefault(&b_info->u.hvm.usb,                false);
>          libxl_defbool_setdefault(&b_info->u.hvm.xen_platform_pci,   true);
> +        libxl_defbool_setdefault(&b_info->u.hvm.ahci,               false);
>  
>          libxl_defbool_setdefault(&b_info->u.hvm.spice.enable, false);
>          if (!libxl_defbool_val(b_info->u.hvm.spice.enable) &&
> diff --git a/tools/libxl/libxl_dm.c b/tools/libxl/libxl_dm.c
> index 33f9ce6..9216028 100644
> --- a/tools/libxl/libxl_dm.c
> +++ b/tools/libxl/libxl_dm.c
> @@ -818,6 +818,8 @@ static int libxl__build_device_model_args_new(libxl__gc 
> *gc,
>      flexarray_append(dm_args, libxl__sprintf(gc, "%"PRId64, ram_size));
>  
>      if (b_info->type == LIBXL_DOMAIN_TYPE_HVM) {
> +        if (libxl_defbool_val(b_info->u.hvm.ahci))
> +            flexarray_append_pair(dm_args, "-device", "ahci,id=ahci0");

Do we still want this even if num_disks == 0?

>          for (i = 0; i < num_disks; i++) {
>              int disk, part;
>              int dev_number =
> @@ -872,7 +874,13 @@ static int libxl__build_device_model_args_new(libxl__gc 
> *gc,
>                      drive = libxl__sprintf
>                          (gc, 
> "file=%s,if=scsi,bus=0,unit=%d,format=%s,cache=writeback",
>                           pdev_path, disk, format);
> -                else if (disk < 4)
> +                else if (disk < 6 && libxl_defbool_val(b_info->u.hvm.ahci)) {
> +                    flexarray_vappend(dm_args, "-drive",
> +                        
> GCSPRINTF("file=%s,if=none,id=ahcidisk-%d,format=%s,cache=writeback",
> +                        pdev_path, disk, format), "-device", 
> GCSPRINTF("ide-hd,bus=ahci0.%d,unit=0,drive=ahcidisk-%d",

This line looks to be overlong, and in general the wrapping in this new
code seems rather arbitrary. 

How about:
                    flexarray_vappend(dm_args,
                        "-drive",
                        
GCSPRINTF("file=%s,if=none,id=ahcidisk-%d,format=%s,cache=writeback",
                                  pdev_path, disk, format),
                        "-device",
                        
GCSPRINTF("ide-hd,bus=ahci0.%d,unit=0,drive=ahcidisk-%d",
                                  disk, disk),
                       NULL);
?

> +                    continue;

From a flow of control PoV this continue is a little jarring.

How about adding a new const char *device = NULL to complement the
existing drive and doing the add in the common bit with an if (device)?



_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel