From: "Jan Beulich" <JBeulich@suse.com>
To: Boris Ostrovsky <boris.ostrovsky@oracle.com>
Cc: tim@xen.org, kevin.tian@intel.com, keir@xen.org,
suravee.suthikulpanit@amd.com, andrew.cooper3@citrix.com,
eddie.dong@intel.com, xen-devel@lists.xen.org,
Aravind.Gopalakrishnan@amd.com, jun.nakajima@intel.com
Subject: Re: [PATCH v10 09/20] x86/VPMU: Add public xenpmu.h
Date: Thu, 11 Sep 2014 15:55:43 +0100 [thread overview]
Message-ID: <5411D40F0200007800033FDA@mail.emea.novell.com> (raw)
In-Reply-To: <5411A97C.7090100@oracle.com>
>>> On 11.09.14 at 15:54, <boris.ostrovsky@oracle.com> wrote:
> On 09/11/2014 02:39 AM, Jan Beulich wrote:
>>>>> On 10.09.14 at 19:23, <boris.ostrovsky@oracle.com> wrote:
>>> On 09/10/2014 10:45 AM, Jan Beulich wrote:
>>>>>>> On 04.09.14 at 05:41, <boris.ostrovsky@oracle.com> wrote:
>>>>> +struct xen_pmu_arch {
>>>>> + union {
>>>>> + struct cpu_user_regs regs;
>>>>> + uint8_t pad[256];
>>>>> + } r;
>>>> Can you remind me again what you need the union and padding for
>>>> here?
>>> This structure is laid out in a shared page with a (possibly 32-bit)
>>> guest who need to access fields that follow this union.
>> Hmm, okay. But how would such a guest make reasonable use of
>> the regs field then?
>
> When hypervisor is preparing this data for 32-bit consumer in
> vpmu_do_interrupts() it translates registers to 32-bit version:
>
> struct compat_cpu_user_regs *cmp;
> gregs = guest_cpu_user_regs();
> cmp = (void *)&vpmu->xenpmu_data->pmu.r.regs;
> XLAT_cpu_user_regs(cmp, gregs);
>
> I remember struggling trying to figure a better way of presenting this
> but ended up with the (void *) cast. IIRC I tried putting
> compat_cpu_user_regs into the union but something didn't quite work
> (with compilation).
Of course that can't work - the compat structure simply doesn't
exist for public headers.
>> And then - why 256 and not 200? struct
>> cpu_user_regs can't change size anyway. Plus, finally, why do
>> you expose the GPRs but not any of the other register state?
>
> I wanted to leave some padding in case we decide to add non-GPR
> registers and keep major version of the interface unchanged (only minor
> version will bumped). TBH though, I can't think of any non-GPR registers
> to be ever useful.
Then what do you need the GPRs for here? I don't think they're
any better or worse than, say, XMM ones. I could see you needing/
wanting some basic stuff like CS:RIP and SS:RSP and maybe EFLAGS,
but that's about it.
Jan
next prev parent reply other threads:[~2014-09-11 14:55 UTC|newest]
Thread overview: 72+ messages / expand[flat|nested] mbox.gz Atom feed top
2014-09-04 3:41 [PATCH v10 00/20] x86/PMU: Xen PMU PV(H) support Boris Ostrovsky
2014-09-04 3:41 ` [PATCH v10 01/20] common/symbols: Export hypervisor symbols to privileged guest Boris Ostrovsky
2014-09-04 3:41 ` [PATCH v10 02/20] x86/VPMU: Manage VPMU_CONTEXT_SAVE flag in vpmu_save_force() Boris Ostrovsky
2014-09-04 3:41 ` [PATCH v10 03/20] x86/VPMU: Set MSR bitmaps only for HVM/PVH guests Boris Ostrovsky
2014-09-04 3:41 ` [PATCH v10 04/20] x86/VPMU: Make vpmu macros a bit more efficient Boris Ostrovsky
2014-09-04 3:41 ` [PATCH v10 05/20] intel/VPMU: Clean up Intel VPMU code Boris Ostrovsky
2014-09-04 3:41 ` [PATCH v10 06/20] vmx: Merge MSR management routines Boris Ostrovsky
2014-09-08 16:07 ` Jan Beulich
2014-09-08 17:28 ` Boris Ostrovsky
2014-09-09 9:11 ` Jan Beulich
2014-09-04 3:41 ` [PATCH v10 07/20] x86/VPMU: Handle APIC_LVTPC accesses Boris Ostrovsky
2014-09-04 3:41 ` [PATCH v10 08/20] intel/VPMU: MSR_CORE_PERF_GLOBAL_CTRL should be initialized to zero Boris Ostrovsky
2014-09-04 3:41 ` [PATCH v10 09/20] x86/VPMU: Add public xenpmu.h Boris Ostrovsky
2014-09-10 14:45 ` Jan Beulich
2014-09-10 17:23 ` Boris Ostrovsky
2014-09-11 6:39 ` Jan Beulich
2014-09-11 13:54 ` Boris Ostrovsky
2014-09-11 14:55 ` Jan Beulich [this message]
2014-09-11 15:26 ` Boris Ostrovsky
2014-09-11 15:59 ` Jan Beulich
2014-09-11 16:51 ` Boris Ostrovsky
2014-09-12 6:50 ` Jan Beulich
2014-09-12 14:21 ` Boris Ostrovsky
2014-09-12 14:38 ` Jan Beulich
2014-09-12 15:18 ` Boris Ostrovsky
2014-09-15 11:56 ` Konrad Rzeszutek Wilk
2014-09-15 13:06 ` Jan Beulich
2014-09-16 1:00 ` Boris Ostrovsky
2014-09-16 0:49 ` Boris Ostrovsky
2014-09-04 3:41 ` [PATCH v10 10/20] x86/VPMU: Make vpmu not HVM-specific Boris Ostrovsky
2014-09-04 3:41 ` [PATCH v10 11/20] x86/VPMU: Interface for setting PMU mode and flags Boris Ostrovsky
2014-09-10 15:05 ` Jan Beulich
2014-09-10 17:37 ` Boris Ostrovsky
2014-09-11 6:44 ` Jan Beulich
2014-09-11 14:12 ` Boris Ostrovsky
2014-09-11 14:59 ` Jan Beulich
2014-09-11 16:10 ` Boris Ostrovsky
2014-09-12 6:49 ` Jan Beulich
2014-09-12 14:12 ` Boris Ostrovsky
2014-09-12 14:39 ` Jan Beulich
2014-09-12 15:03 ` Boris Ostrovsky
2014-09-12 15:30 ` Jan Beulich
2014-09-12 15:54 ` Boris Ostrovsky
2014-09-12 16:05 ` Jan Beulich
2014-09-12 11:41 ` Dietmar Hahn
2014-09-12 14:25 ` Boris Ostrovsky
2014-09-15 13:35 ` Dietmar Hahn
2014-09-18 4:11 ` Tian, Kevin
2014-09-18 21:50 ` Boris Ostrovsky
2014-09-19 6:51 ` Jan Beulich
2014-09-19 12:42 ` Boris Ostrovsky
2014-09-19 13:28 ` Jan Beulich
2014-09-22 22:29 ` Tian, Kevin
2014-09-22 22:32 ` Tian, Kevin
2014-09-22 22:48 ` Boris Ostrovsky
2014-09-04 3:41 ` [PATCH v10 12/20] x86/VPMU: Initialize PMU for PV(H) guests Boris Ostrovsky
2014-09-04 3:41 ` [PATCH v10 13/20] x86/VPMU: When handling MSR accesses, leave fault injection to callers Boris Ostrovsky
2014-09-18 5:01 ` Tian, Kevin
2014-09-04 3:41 ` [PATCH v10 14/20] x86/VPMU: Add support for PMU register handling on PV guests Boris Ostrovsky
2014-09-04 3:41 ` [PATCH v10 15/20] x86/VPMU: Handle PMU interrupts for " Boris Ostrovsky
2014-09-10 15:30 ` Jan Beulich
2014-09-04 3:41 ` [PATCH v10 16/20] x86/VPMU: Merge vpmu_rdmsr and vpmu_wrmsr Boris Ostrovsky
2014-09-10 15:33 ` Jan Beulich
2014-09-18 4:16 ` Tian, Kevin
2014-09-04 3:41 ` [PATCH v10 17/20] x86/VPMU: Add privileged PMU mode Boris Ostrovsky
2014-09-10 15:39 ` Jan Beulich
2014-09-04 3:41 ` [PATCH v10 18/20] x86/VPMU: Save VPMU state for PV guests during context switch Boris Ostrovsky
2014-09-10 15:44 ` Jan Beulich
2014-09-04 3:41 ` [PATCH v10 19/20] x86/VPMU: NMI-based VPMU support Boris Ostrovsky
2014-09-04 3:41 ` [PATCH v10 20/20] x86/VPMU: Move VPMU files up from hvm/ directory Boris Ostrovsky
2014-09-10 15:48 ` Jan Beulich
2014-09-10 15:54 ` [PATCH v10 00/20] x86/PMU: Xen PMU PV(H) support Jan Beulich
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=5411D40F0200007800033FDA@mail.emea.novell.com \
--to=jbeulich@suse.com \
--cc=Aravind.Gopalakrishnan@amd.com \
--cc=andrew.cooper3@citrix.com \
--cc=boris.ostrovsky@oracle.com \
--cc=eddie.dong@intel.com \
--cc=jun.nakajima@intel.com \
--cc=keir@xen.org \
--cc=kevin.tian@intel.com \
--cc=suravee.suthikulpanit@amd.com \
--cc=tim@xen.org \
--cc=xen-devel@lists.xen.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).