xen-devel.lists.xenproject.org archive mirror
 help / color / mirror / Atom feed
From: George Dunlap <george.dunlap@eu.citrix.com>
To: David Vrabel <david.vrabel@citrix.com>
Cc: Frediano Ziglio <frediano.ziglio@citrix.com>,
	"xen-devel@lists.xensource.com" <xen-devel@lists.xensource.com>
Subject: Re: [PATCH 1/3] trace: allow for different sub-classes of TRC_PV_* tracepoints
Date: Thu, 31 May 2012 15:13:57 +0100	[thread overview]
Message-ID: <4FC77CA5.1010308@eu.citrix.com> (raw)
In-Reply-To: <1338462397-32111-2-git-send-email-david.vrabel@citrix.com>

On 31/05/12 12:06, David Vrabel wrote:
> From: David Vrabel<david.vrabel@citrix.com>
>
> We want to add additional sub-classes for TRC_PV tracepoints and to be
> able to only capture these new sub-classes.  This cannot currently be
> done as the existing tracepoints all use a sub-class of 0xf.
>
> So, redefine the PV events to use a new sub-class.  All the current
> tracepoints are tracing entry points to the hypervisor so the
> sub-class is named TRC_PV_ENTRY.
>
> This change does not affect xenalyze as that only looks at the main
> class and the event number and does not use the sub-class field.
>
> Signed-off-by: Frediano Ziglio<frediano.ziglio@citrix.com>
> Signed-off-by: David Vrabel<david.vrabel@citrix.com>
As post-4.2 material:

Acked-by: George Dunlap <george.dunlap@eu.citrix.com>
> ---
>   tools/xentrace/formats     |   44 ++++++++++++++++++++++----------------------
>   xen/include/public/trace.h |   35 +++++++++++++++++++++--------------
>   2 files changed, 43 insertions(+), 36 deletions(-)
>
> diff --git a/tools/xentrace/formats b/tools/xentrace/formats
> index 04e54d5..71220c0 100644
> --- a/tools/xentrace/formats
> +++ b/tools/xentrace/formats
> @@ -82,28 +82,28 @@
>   0x0010f002  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  page_grant_unmap    [ domid = %(1)d ]
>   0x0010f003  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  page_grant_transfer [ domid = %(1)d ]
>
> -0x0020f001  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  hypercall  [ eip = 0x%(1)08x, eax = 0x%(2)08x ]
> -0x0020f101  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  hypercall  [ rip = 0x%(2)08x%(1)08x, eax = 0x%(3)08x ]
> -0x0020f003  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  trap       [ eip = 0x%(1)08x, trapnr:error = 0x%(2)08x ]
> -0x0020f103  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  trap       [ rip = 0x%(2)08x%(1)08x, trapnr:error = 0x%(3)08x ]
> -0x0020f004  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  page_fault [ eip = 0x%(1)08x, addr = 0x%(2)08x, error = 0x%(3)08x ]
> -0x0020f104  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  page_fault [ rip = 0x%(2)08x%(1)08x, addr = 0x%(4)08x%(3)08x, error = 0x%(5)08x ]
> -0x0020f005  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  forced_invalid_op   [ eip = 0x%(1)08x ]
> -0x0020f105  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  forced_invalid_op   [ rip = 0x%(2)08x%(1)08x ]
> -0x0020f006  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  emulate_privop      [ eip = 0x%(1)08x ]
> -0x0020f106  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  emulate_privop      [ rip = 0x%(2)08x%(1)08x ]
> -0x0020f007  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  emulate_4G          [ eip = 0x%(1)08x ]
> -0x0020f107  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  emulate_4G          [ rip = 0x%(2)08x%(1)08x ]
> -0x0020f008  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  math_state_restore
> -0x0020f108  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  math_state_restore
> -0x0020f009  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  paging_fixup        [ eip = 0x%(1)08x, addr = 0x%(2)08x ]
> -0x0020f109  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  paging_fixup        [ rip = 0x%(2)08x%(1)08x, addr = 0x%(4)08x%(3)08x ]
> -0x0020f00a  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  gdt_ldt_mapping_fault  [ eip = 0x%(1)08x, offset = 0x%(2)08x ]
> -0x0020f10a  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  gdt_ldt_mapping_fault  [ rip = 0x%(2)08x%(1)08x, offset = 0x%(4)08x%(3)08x ]
> -0x0020f00b  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  ptwr_emulation      [ addr = 0x%(3)08x, eip = 0x%(4)08x, npte = 0x%(2)08x%(1)08x ]
> -0x0020f10b  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  ptwr_emulation      [ addr = 0x%(4)08x%(3)08x, rip = 0x%(6)08x%(5)08x, npte = 0x%(2)08x%(1)08x ]
> -0x0020f00c  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  ptwr_emulation_pae  [ addr = 0x%(3)08x, eip = 0x%(4)08x, npte = 0x%(2)08x%(1)08x ]
> -0x0020f10c  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  ptwr_emulation_pae  [ addr = 0x%(4)08x%(3)08x, rip = 0x%(6)08x%(5)08x, npte = 0x%(2)08x%(1)08x ]
> +0x00201001  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  hypercall  [ eip = 0x%(1)08x, eax = 0x%(2)08x ]
> +0x00201101  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  hypercall  [ rip = 0x%(2)08x%(1)08x, eax = 0x%(3)08x ]
> +0x00201003  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  trap       [ eip = 0x%(1)08x, trapnr:error = 0x%(2)08x ]
> +0x00201103  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  trap       [ rip = 0x%(2)08x%(1)08x, trapnr:error = 0x%(3)08x ]
> +0x00201004  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  page_fault [ eip = 0x%(1)08x, addr = 0x%(2)08x, error = 0x%(3)08x ]
> +0x00201104  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  page_fault [ rip = 0x%(2)08x%(1)08x, addr = 0x%(4)08x%(3)08x, error = 0x%(5)08x ]
> +0x00201005  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  forced_invalid_op   [ eip = 0x%(1)08x ]
> +0x00201105  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  forced_invalid_op   [ rip = 0x%(2)08x%(1)08x ]
> +0x00201006  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  emulate_privop      [ eip = 0x%(1)08x ]
> +0x00201106  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  emulate_privop      [ rip = 0x%(2)08x%(1)08x ]
> +0x00201007  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  emulate_4G          [ eip = 0x%(1)08x ]
> +0x00201107  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  emulate_4G          [ rip = 0x%(2)08x%(1)08x ]
> +0x00201008  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  math_state_restore
> +0x00201108  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  math_state_restore
> +0x00201009  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  paging_fixup        [ eip = 0x%(1)08x, addr = 0x%(2)08x ]
> +0x00201109  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  paging_fixup        [ rip = 0x%(2)08x%(1)08x, addr = 0x%(4)08x%(3)08x ]
> +0x0020100a  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  gdt_ldt_mapping_fault  [ eip = 0x%(1)08x, offset = 0x%(2)08x ]
> +0x0020110a  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  gdt_ldt_mapping_fault  [ rip = 0x%(2)08x%(1)08x, offset = 0x%(4)08x%(3)08x ]
> +0x0020100b  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  ptwr_emulation      [ addr = 0x%(3)08x, eip = 0x%(4)08x, npte = 0x%(2)08x%(1)08x ]
> +0x0020110b  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  ptwr_emulation      [ addr = 0x%(4)08x%(3)08x, rip = 0x%(6)08x%(5)08x, npte = 0x%(2)08x%(1)08x ]
> +0x0020100c  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  ptwr_emulation_pae  [ addr = 0x%(3)08x, eip = 0x%(4)08x, npte = 0x%(2)08x%(1)08x ]
> +0x0020110c  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  ptwr_emulation_pae  [ addr = 0x%(4)08x%(3)08x, rip = 0x%(6)08x%(5)08x, npte = 0x%(2)08x%(1)08x ]
>
>   0x0040f001  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  shadow_not_shadow                 [ gl1e = 0x%(2)08x%(1)08x, va = 0x%(3)08x, flags = 0x%(4)08x ]
>   0x0040f101  CPU%(cpu)d  %(tsc)d (+%(reltsc)8d)  shadow_not_shadow                 [ gl1e = 0x%(2)08x%(1)08x, va = 0x%(4)08x%(3)08x, flags = 0x%(5)08x ]
> diff --git a/xen/include/public/trace.h b/xen/include/public/trace.h
> index 0dfabe9..1f154bb 100644
> --- a/xen/include/public/trace.h
> +++ b/xen/include/public/trace.h
> @@ -94,20 +94,19 @@
>   #define TRC_MEM_POD_ZERO_RECLAIM    (TRC_MEM + 17)
>   #define TRC_MEM_POD_SUPERPAGE_SPLINTER (TRC_MEM + 18)
>
> -
> -#define TRC_PV_HYPERCALL             (TRC_PV +  1)
> -#define TRC_PV_TRAP                  (TRC_PV +  3)
> -#define TRC_PV_PAGE_FAULT            (TRC_PV +  4)
> -#define TRC_PV_FORCED_INVALID_OP     (TRC_PV +  5)
> -#define TRC_PV_EMULATE_PRIVOP        (TRC_PV +  6)
> -#define TRC_PV_EMULATE_4GB           (TRC_PV +  7)
> -#define TRC_PV_MATH_STATE_RESTORE    (TRC_PV +  8)
> -#define TRC_PV_PAGING_FIXUP          (TRC_PV +  9)
> -#define TRC_PV_GDT_LDT_MAPPING_FAULT (TRC_PV + 10)
> -#define TRC_PV_PTWR_EMULATION        (TRC_PV + 11)
> -#define TRC_PV_PTWR_EMULATION_PAE    (TRC_PV + 12)
> -  /* Indicates that addresses in trace record are 64 bits */
> -#define TRC_64_FLAG               (0x100)
> +#define TRC_PV_ENTRY 0x00201000 /* Hypervisor entry points for PV guests. */
> +
> +#define TRC_PV_HYPERCALL             (TRC_PV_ENTRY +  1)
> +#define TRC_PV_TRAP                  (TRC_PV_ENTRY +  3)
> +#define TRC_PV_PAGE_FAULT            (TRC_PV_ENTRY +  4)
> +#define TRC_PV_FORCED_INVALID_OP     (TRC_PV_ENTRY +  5)
> +#define TRC_PV_EMULATE_PRIVOP        (TRC_PV_ENTRY +  6)
> +#define TRC_PV_EMULATE_4GB           (TRC_PV_ENTRY +  7)
> +#define TRC_PV_MATH_STATE_RESTORE    (TRC_PV_ENTRY +  8)
> +#define TRC_PV_PAGING_FIXUP          (TRC_PV_ENTRY +  9)
> +#define TRC_PV_GDT_LDT_MAPPING_FAULT (TRC_PV_ENTRY + 10)
> +#define TRC_PV_PTWR_EMULATION        (TRC_PV_ENTRY + 11)
> +#define TRC_PV_PTWR_EMULATION_PAE    (TRC_PV_ENTRY + 12)
>
>   #define TRC_SHADOW_NOT_SHADOW                 (TRC_SHADOW +  1)
>   #define TRC_SHADOW_FAST_PROPAGATE             (TRC_SHADOW +  2)
> @@ -187,6 +186,14 @@
>   #define TRC_HW_IRQ_UNMAPPED_VECTOR    (TRC_HW_IRQ + 0x7)
>   #define TRC_HW_IRQ_HANDLED            (TRC_HW_IRQ + 0x8)
>
> +/*
> + * Event Flags
> + *
> + * Some events (e.g, TRC_PV_TRAP and TRC_HVM_IOMEM_READ) have multiple
> + * record formats.  These event flags distinguish between the
> + * different formats.
> + */
> +#define TRC_64_FLAG 0x100 /* Addresses are 64 bits (instead of 32 bits) */
>
>   /* This structure represents a single trace buffer record. */
>   struct t_rec {

  reply	other threads:[~2012-05-31 14:13 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-05-31 11:06 [PATCHv3 0/3] trace: improve hypercall tracing David Vrabel
2012-05-31 11:06 ` [PATCH 1/3] trace: allow for different sub-classes of TRC_PV_* tracepoints David Vrabel
2012-05-31 14:13   ` George Dunlap [this message]
2012-05-31 11:06 ` [PATCH 2/3] trace: improve usefulness of hypercall trace record David Vrabel
2012-05-31 14:14   ` George Dunlap
2012-05-31 11:06 ` [PATCH 3/3] trace: trace hypercalls inside a multicall David Vrabel
2012-05-31 14:15   ` George Dunlap
2012-05-31 14:21 ` [PATCHv3 0/3] trace: improve hypercall tracing George Dunlap
2012-05-31 14:32   ` David Vrabel
2012-06-01 16:17     ` George Dunlap
  -- strict thread matches above, loose matches on Subject: below --
2012-10-01 17:47 [PATCHv4 " David Vrabel
2012-10-01 17:47 ` [PATCH 1/3] trace: allow for different sub-classes of TRC_PV_* tracepoints David Vrabel
2012-10-02 15:56   ` George Dunlap

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=4FC77CA5.1010308@eu.citrix.com \
    --to=george.dunlap@eu.citrix.com \
    --cc=david.vrabel@citrix.com \
    --cc=frediano.ziglio@citrix.com \
    --cc=xen-devel@lists.xensource.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).