From: Alexander Graf <agraf@suse.de>
To: David Gibson <david@gibson.dropbear.id.au>
Cc: qemu-devel@nongnu.org
Subject: Re: [Qemu-devel] [PATCH 2/2] pseries: Add device tree properties for VMX/VSX and DFP under kvm
Date: Wed, 12 Oct 2011 16:17:56 +0200 [thread overview]
Message-ID: <4E95A194.5030608@suse.de> (raw)
In-Reply-To: <1318307461-9817-2-git-send-email-david@gibson.dropbear.id.au>
On 10/11/2011 06:31 AM, David Gibson wrote:
> Sufficiently recent PAPR specifications define properties "ibm,vmx"
> and "ibm,dfp" on the CPU node which advertise whether the VMX vector
> extensions (or the later VSX version) and/or the Decimal Floating
> Point operations from IBM's recent POWER CPUs are available.
>
> Currently we do not put these in the guest device tree and the guest
> kernel will consequently assume they are not available. This is good,
> because they are not supported under TCG. VMX is similar enough to
> Altivec that it might be trivial to support, but VSX and DFP would
> both require significant work to support in TCG.
>
> However, when running under kvm on a host which supports these
> instructions, there's no reason not to let the guest use them. This
> patch, therefore, checks for the relevant support on the host CPU
> and, if present, advertises them to the guest as well.
>
> Signed-off-by: David Gibson<david@gibson.dropbear.id.au>
> ---
> hw/spapr.c | 17 +++++++++++++++++
> target-ppc/kvm.c | 10 ++++++++++
> target-ppc/kvm_ppc.h | 12 ++++++++++++
> 3 files changed, 39 insertions(+), 0 deletions(-)
>
> diff --git a/hw/spapr.c b/hw/spapr.c
> index 9a3a1ea..00b9c67 100644
> --- a/hw/spapr.c
> +++ b/hw/spapr.c
> @@ -186,6 +186,8 @@ static void *spapr_create_fdt_skel(const char *cpu_model,
> 0xffffffff, 0xffffffff};
> uint32_t tbfreq = kvm_enabled() ? kvmppc_get_tbfreq() : TIMEBASE_FREQ;
> uint32_t cpufreq = kvm_enabled() ? kvmppc_get_clockfreq() : 1000000000;
> + uint32_t vmx = kvm_enabled() ? kvmppc_get_vmx() : 0;
> + uint32_t dfp = kvm_enabled() ? kvmppc_get_dfp() : 0;
This should only happen when the CPU type selected by -cpu also supports
VMX or DFP. For now, this is moot as we can't run any non-host CPU in
KVM for HV PPC. But looking forward, it certainly makes sense to split it.
This is not a nack, but please make sure to implement -cpu host in the
near future and default to that for -M pseries. Then we can change this
piece here to check for vcpu capabilities and AND them with the host caps.
The rest looks good.
Alex
>
> if ((index % smt) != 0) {
> continue;
> @@ -233,6 +235,21 @@ static void *spapr_create_fdt_skel(const char *cpu_model,
> segs, sizeof(segs))));
> }
>
> + /* Advertise VMX/VSX (vector extensions) if available
> + * 0 / no property == no vector extensions
> + * 1 == VMX / Altivec available
> + * 2 == VSX available */
> + if (vmx) {
> + _FDT((fdt_property_cell(fdt, "ibm,vmx", vmx)));
> + }
> +
> + /* Advertise DFP (Decimal Floating Point) if available
> + * 0 / no property == no DFP
> + * 1 == DFP available */
> + if (dfp) {
> + _FDT((fdt_property_cell(fdt, "ibm,dfp", dfp)));
> + }
> +
> _FDT((fdt_end_node(fdt)));
> }
>
> diff --git a/target-ppc/kvm.c b/target-ppc/kvm.c
> index 6667b61..6a48eb4 100644
> --- a/target-ppc/kvm.c
> +++ b/target-ppc/kvm.c
> @@ -694,6 +694,16 @@ uint64_t kvmppc_get_clockfreq(void)
> return kvmppc_read_int_cpu_dt("clock-frequency");
> }
>
> +uint32_t kvmppc_get_vmx(void)
> +{
> + return kvmppc_read_int_cpu_dt("ibm,vmx");
> +}
> +
> +uint32_t kvmppc_get_dfp(void)
> +{
> + return kvmppc_read_int_cpu_dt("ibm,dfp");
> +}
> +
> int kvmppc_get_hypercall(CPUState *env, uint8_t *buf, int buf_len)
> {
> uint32_t *hc = (uint32_t*)buf;
> diff --git a/target-ppc/kvm_ppc.h b/target-ppc/kvm_ppc.h
> index 9e8a7b5..fa131bf 100644
> --- a/target-ppc/kvm_ppc.h
> +++ b/target-ppc/kvm_ppc.h
> @@ -15,6 +15,8 @@ void kvmppc_init(void);
>
> uint32_t kvmppc_get_tbfreq(void);
> uint64_t kvmppc_get_clockfreq(void);
> +uint32_t kvmppc_get_vmx(void);
> +uint32_t kvmppc_get_dfp(void);
> int kvmppc_get_hypercall(CPUState *env, uint8_t *buf, int buf_len);
> int kvmppc_set_interrupt(CPUState *env, int irq, int level);
> void kvmppc_set_papr(CPUState *env);
> @@ -35,6 +37,16 @@ static inline uint64_t kvmppc_get_clockfreq(void)
> return 0;
> }
>
> +static inline uint32_t kvmppc_get_vmx(void)
> +{
> + return 0;
> +}
> +
> +static inline uint32_t kvmppc_get_dfp(void)
> +{
> + return 0;
> +}
> +
> static inline int kvmppc_get_hypercall(CPUState *env, uint8_t *buf, int buf_len)
> {
> return -1;
next prev parent reply other threads:[~2011-10-12 14:15 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2011-10-11 4:31 [Qemu-devel] [PATCH 1/2] ppc: Generalize the kvmppc_get_clockfreq() function David Gibson
2011-10-11 4:31 ` [Qemu-devel] [PATCH 2/2] pseries: Add device tree properties for VMX/VSX and DFP under kvm David Gibson
2011-10-12 14:17 ` Alexander Graf [this message]
2011-10-12 14:19 ` [Qemu-devel] [PATCH 1/2] ppc: Generalize the kvmppc_get_clockfreq() function Alexander Graf
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=4E95A194.5030608@suse.de \
--to=agraf@suse.de \
--cc=david@gibson.dropbear.id.au \
--cc=qemu-devel@nongnu.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).