linux-kernel.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com>
To: Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com>
Cc: Avi Kivity <avi@redhat.com>,
	Marcelo Tosatti <mtosatti@redhat.com>,
	LKML <linux-kernel@vger.kernel.org>, KVM <kvm@vger.kernel.org>
Subject: [PATCH 12/13] KVM: MMU: trace fast page fault
Date: Thu, 29 Mar 2012 17:27:31 +0800	[thread overview]
Message-ID: <4F742B03.3040800@linux.vnet.ibm.com> (raw)
In-Reply-To: <4F742951.7080003@linux.vnet.ibm.com>

To see what happen on this path and help us to optimize it

Signed-off-by: Xiao Guangrong <xiaoguangrong@linux.vnet.ibm.com>
---
 arch/x86/kvm/mmu.c      |    2 ++
 arch/x86/kvm/mmutrace.h |   43 +++++++++++++++++++++++++++++++++++++++++++
 2 files changed, 45 insertions(+), 0 deletions(-)

diff --git a/arch/x86/kvm/mmu.c b/arch/x86/kvm/mmu.c
index 4a01be4..94cd31d 100644
--- a/arch/x86/kvm/mmu.c
+++ b/arch/x86/kvm/mmu.c
@@ -3011,6 +3011,8 @@ fast_page_fault(struct kvm_vcpu *vcpu, gva_t gva, gfn_t gfn, int level,
 				       expected_access, fn);

 exit:
+	trace_fast_page_fault(vcpu, gva, gfn, error_code, iterator.sptep,
+			      spte, ret);
 	walk_shadow_page_lockless_end(vcpu);

 	return ret;
diff --git a/arch/x86/kvm/mmutrace.h b/arch/x86/kvm/mmutrace.h
index 89fb0e8..a81c8a7 100644
--- a/arch/x86/kvm/mmutrace.h
+++ b/arch/x86/kvm/mmutrace.h
@@ -243,6 +243,49 @@ TRACE_EVENT(
 	TP_printk("addr:%llx gfn %llx access %x", __entry->addr, __entry->gfn,
 		  __entry->access)
 );
+
+static u32 page_fault_expected_access(u32 error_code);
+static bool spte_satisfied(u64 spte, u32 access);
+#define __spte_satisfied(__spte)				\
+	(__entry->success && spte_satisfied(__entry->__spte,	\
+	      page_fault_expected_access(__entry->error_code)))
+TRACE_EVENT(
+	fast_page_fault,
+	TP_PROTO(struct kvm_vcpu *vcpu, gva_t gva, gfn_t gfn, u32 error_code,
+		 u64 *sptep, u64 old_spte, bool retry),
+	TP_ARGS(vcpu, gva, gfn, error_code, sptep, old_spte, retry),
+
+	TP_STRUCT__entry(
+		__field(int, vcpu_id)
+		__field(gva_t, gva)
+		__field(gfn_t, gfn)
+		__field(u32, error_code)
+		__field(u64 *, sptep)
+		__field(u64, old_spte)
+		__field(u64, new_spte)
+		__field(bool, success)
+	),
+
+	TP_fast_assign(
+		__entry->vcpu_id = vcpu->vcpu_id;
+		__entry->gva = gva;
+		__entry->gfn = gfn;
+		__entry->error_code = error_code;
+		__entry->sptep = sptep;
+		__entry->old_spte = old_spte;
+		__entry->new_spte = *sptep;
+		__entry->success = retry;
+	),
+
+	TP_printk("vcpu %d gva %lx gfn %llx error_code %s sptep %p "
+		  "old %#llx new %llx spurious %d fixed %d", __entry->vcpu_id,
+		  __entry->gva, __entry->gfn,
+		  __print_flags(__entry->error_code, "|",
+				kvm_mmu_trace_pferr_flags),
+		  __entry->sptep, __entry->old_spte, __entry->new_spte,
+		  __spte_satisfied(old_spte), __spte_satisfied(new_spte)
+	)
+);
 #endif /* _TRACE_KVMMMU_H */

 #undef TRACE_INCLUDE_PATH
-- 
1.7.7.6


  parent reply	other threads:[~2012-03-29  9:27 UTC|newest]

Thread overview: 83+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-03-29  9:20 [PATCH 00/13] KVM: MMU: fast page fault Xiao Guangrong
2012-03-29  9:20 ` [PATCH 01/13] KVM: MMU: properly assert spte on rmap_next path Xiao Guangrong
2012-03-29  9:21 ` [PATCH 02/13] KVM: MMU: abstract spte write-protect Xiao Guangrong
2012-03-29 11:11   ` Avi Kivity
2012-03-29 11:51     ` Xiao Guangrong
2012-03-29  9:22 ` [PATCH 03/13] KVM: MMU: split FNAME(prefetch_invalid_gpte) Xiao Guangrong
2012-03-29 13:00   ` Avi Kivity
2012-03-30  3:51     ` Xiao Guangrong
2012-03-29  9:22 ` [PATCH 04/13] KVM: MMU: introduce FNAME(get_sp_gpa) Xiao Guangrong
2012-03-29 13:07   ` Avi Kivity
2012-03-30  5:01     ` Xiao Guangrong
2012-04-01 12:42       ` Avi Kivity
2012-03-29  9:23 ` [PATCH 05/13] KVM: MMU: reset shadow_mmio_mask Xiao Guangrong
2012-03-29 13:10   ` Avi Kivity
2012-03-29 15:28     ` Avi Kivity
2012-03-29 16:24       ` Avi Kivity
2012-03-29  9:23 ` [PATCH 06/13] KVM: VMX: export PFEC.P bit on ept Xiao Guangrong
2012-03-29  9:24 ` [PATCH 07/13] KVM: MMU: store more bits in rmap Xiao Guangrong
2012-03-29  9:25 ` [PATCH 08/13] KVM: MMU: fask check whether page is writable Xiao Guangrong
2012-03-29 15:49   ` Avi Kivity
2012-03-30  5:10     ` Xiao Guangrong
2012-04-01 15:52   ` Avi Kivity
2012-04-05 17:54     ` Xiao Guangrong
2012-04-12 23:08       ` Marcelo Tosatti
2012-04-13 10:26         ` Xiao Guangrong
2012-03-29  9:25 ` [PATCH 09/13] KVM: MMU: get expected spte out of mmu-lock Xiao Guangrong
2012-04-01 15:53   ` Avi Kivity
2012-04-05 18:25     ` Xiao Guangrong
2012-04-09 12:28       ` Avi Kivity
2012-04-09 13:16         ` Takuya Yoshikawa
2012-04-09 13:21           ` Avi Kivity
2012-03-29  9:26 ` [PATCH 10/13] KVM: MMU: store vcpu id in spte to notify page write-protect path Xiao Guangrong
2012-03-29  9:27 ` [PATCH 11/13] KVM: MMU: fast path of handling guest page fault Xiao Guangrong
2012-03-31 12:24   ` Xiao Guangrong
2012-04-01 16:23   ` Avi Kivity
2012-04-03 13:04     ` Avi Kivity
2012-04-05 19:39     ` Xiao Guangrong
2012-03-29  9:27 ` Xiao Guangrong [this message]
2012-03-29  9:28 ` [PATCH 13/13] KVM: MMU: fix kvm_mmu_pagetable_walk tracepoint Xiao Guangrong
2012-03-29 10:18 ` [PATCH 00/13] KVM: MMU: fast page fault Avi Kivity
2012-03-29 11:40   ` Xiao Guangrong
2012-03-29 12:57     ` Avi Kivity
2012-03-30  9:18       ` Xiao Guangrong
2012-03-31 13:12         ` Xiao Guangrong
2012-04-01 12:58         ` Avi Kivity
2012-04-05 21:57           ` Xiao Guangrong
2012-04-06  5:24             ` Xiao Guangrong
2012-04-09 13:20               ` Avi Kivity
2012-04-09 13:59                 ` Xiao Guangrong
2012-04-09 13:12 ` Avi Kivity
2012-04-09 13:55   ` Xiao Guangrong
2012-04-09 14:01     ` Xiao Guangrong
2012-04-09 14:25     ` Avi Kivity
2012-04-09 17:58   ` Marcelo Tosatti
2012-04-09 18:13     ` Xiao Guangrong
2012-04-09 19:31       ` Marcelo Tosatti
2012-04-09 18:26     ` Xiao Guangrong
2012-04-09 19:46       ` Marcelo Tosatti
2012-04-10  3:06         ` Xiao Guangrong
2012-04-10 10:04         ` Avi Kivity
2012-04-11  1:47           ` Marcelo Tosatti
2012-04-11  9:15             ` Avi Kivity
2012-04-10 10:39         ` Avi Kivity
2012-04-10 11:40           ` Takuya Yoshikawa
2012-04-10 11:58             ` Xiao Guangrong
2012-04-11 12:15               ` Takuya Yoshikawa
2012-04-11 12:38                 ` Xiao Guangrong
2012-04-11 14:14                   ` Takuya Yoshikawa
2012-04-11 14:21                     ` Avi Kivity
2012-04-11 22:26                       ` Takuya Yoshikawa
2012-04-13 14:25                     ` Takuya Yoshikawa
2012-04-15  9:32                       ` Avi Kivity
2012-04-16 15:49                         ` Takuya Yoshikawa
2012-04-16 16:02                           ` Avi Kivity
2012-04-17  6:26                           ` Xiao Guangrong
2012-04-17  7:51                             ` Avi Kivity
2012-04-17 12:37                               ` Takuya Yoshikawa
2012-04-17 12:41                                 ` Avi Kivity
2012-04-17 14:54                                   ` Takuya Yoshikawa
2012-04-17 14:56                                     ` Avi Kivity
2012-04-18 13:42                                       ` Takuya Yoshikawa
2012-04-17  6:16                         ` Xiao Guangrong
2012-04-10 10:10       ` Avi Kivity

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=4F742B03.3040800@linux.vnet.ibm.com \
    --to=xiaoguangrong@linux.vnet.ibm.com \
    --cc=avi@redhat.com \
    --cc=kvm@vger.kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=mtosatti@redhat.com \
    /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).