On 7/22/2016 4:00 PM, Razvan Cojocaru wrote:
On 07/22/2016 03:32 PM, Corneliu ZUZU wrote:
Look @ hvm_do_resume():
1. If you, as a toolstack user, get at sane step no. 6: "Uninitialize
everything (no events are possible here because of steps 4-5)."
2. But just bef
On 7/22/2016 3:10 PM, Razvan Cojocaru wrote:
On 07/22/2016 02:39 PM, Corneliu ZUZU wrote:
On 7/22/2016 1:29 PM, Razvan Cojocaru wrote:
On 07/22/2016 01:17 PM, Corneliu ZUZU wrote:
On 7/22/2016 12:55 PM, Razvan Cojocaru wrote:
On 07/22/2016 12:27 PM, Corneliu ZUZU wrote:
Hi,
I've been
On 7/22/2016 2:13 PM, Andrew Cooper wrote:
On 22/07/16 11:31, Corneliu ZUZU wrote:
On 7/22/2016 12:51 PM, Andrew Cooper wrote:
On 22/07/16 10:27, Corneliu ZUZU wrote:
Hi,
I've been inspecting vm-event code parts to try and understand when
and why domain pausing/locking is done. If I
On 7/22/2016 1:29 PM, Razvan Cojocaru wrote:
On 07/22/2016 01:17 PM, Corneliu ZUZU wrote:
On 7/22/2016 12:55 PM, Razvan Cojocaru wrote:
On 07/22/2016 12:27 PM, Corneliu ZUZU wrote:
Hi,
I've been inspecting vm-event code parts to try and understand when and
why domain pausing/locking is done
On 7/22/2016 12:51 PM, Andrew Cooper wrote:
On 22/07/16 10:27, Corneliu ZUZU wrote:
Hi,
I've been inspecting vm-event code parts to try and understand when
and why domain pausing/locking is done. If I understood correctly,
domain pausing is done solely to force all the vCPUs of that domain
On 7/22/2016 12:55 PM, Razvan Cojocaru wrote:
On 07/22/2016 12:27 PM, Corneliu ZUZU wrote:
Hi,
I've been inspecting vm-event code parts to try and understand when and
why domain pausing/locking is done. If I understood correctly, domain
pausing is done solely to force all the vCPUs
Hi,
I've been inspecting vm-event code parts to try and understand when and
why domain pausing/locking is done. If I understood correctly, domain
pausing is done solely to force all the vCPUs of that domain to see a
configuration update and act upon it (e.g. in the case of a
On 7/18/2016 9:07 PM, Andrew Cooper wrote:
On 15/07/16 08:18, Corneliu ZUZU wrote:
On 7/12/2016 9:10 AM, Corneliu ZUZU wrote:
On 7/11/2016 7:43 PM, Tamas K Lengyel wrote:
On Sat, Jul 9, 2016 at 12:46 PM, Corneliu ZUZU
<cz...@bitdefender.com> wrote:
On 7/9/2016 9:10 PM, Tamas K Lengyel
On 7/15/2016 4:05 PM, Andrew Cooper wrote:
On 15/07/16 11:41, Corneliu ZUZU wrote:
Corneliu ZUZU (7):
1/7: asm-arm/atomic.h: fix arm32|arm64 macros duplication
Reviewed-by: Stefano Stabellini <sstabell...@kernel.org>
Changed:
* remove paragraph(s
implemented X86 functions in
common
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
Reviewed-by: Andrew Cooper <andrew.coop...@citrix.com>
Acked-by: Julien Grall <julien.gr...@arm.com>
---
Changed since v4:
---
xen/include/asm-arm/atomic.h | 5 +
xen/include/asm
Turn atomic_inc_return and atomic_dec_return atomic.h macros to inline
functions.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
Reviewed-by: Stefano Stabellini <sstabell...@kernel.org>
---
Changed since v4:
* README.LinuxPrimitives not updated anymore
---
xen/include/asm-
This wouldn't let me make a param of a function that used atomic_read() const.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
Reviewed-by: Andrew Cooper <andrew.coop...@citrix.com>
Reviewed-by: Stefano Stabellini <sstabell...@kernel.org>
Reviewed-by: Julien Grall <
.
Removed outdated comment ("NB. I've [...]").
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
Suggested-by: Andrew Cooper <andrew.coop...@citrix.com>
Reviewed-by: Andrew Cooper <andrew.coop...@citrix.com>
Reviewed-by: Stefano Stabellini <sstabell...@kernel.org>
Place atomic_inc_and_test() implementation after atomic_inc().
Also empty line fix.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
Reviewed-by: Andrew Cooper <andrew.coop...@citrix.com>
---
Changed since v4:
---
xen/include/asm-x86/atomic.h | 39 +++---
Move duplicate macros between asm-arm/arm32/atomic.h and asm-arm/arm64/atomic.h
to asm-arm/atomic.h.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
Reviewed-by: Stefano Stabellini <sstabell...@kernel.org>
---
Changed since v4:
* remove paragraph(s) from README.LinuxPri
Corneliu ZUZU (7):
1/7: asm-arm/atomic.h: fix arm32|arm64 macros duplication
Reviewed-by: Stefano Stabellini <sstabell...@kernel.org>
Changed:
* remove paragraph(s) from README.LinuxPrimitives file
* no empty line additions
2/7: asm-x86/atomic.h: minor:
On 7/15/2016 1:06 PM, Julien Grall wrote:
On 15/07/16 10:55, Corneliu ZUZU wrote:
On 7/15/2016 12:28 PM, Julien Grall wrote:
Hi Corneliu,
On 15/07/16 07:48, Corneliu ZUZU wrote:
Move duplicate macros between asm-arm/arm32/atomic.h and
asm-arm/arm64/atomic.h
to asm-arm/atomic.h. Adjust
On 7/15/2016 12:30 PM, Julien Grall wrote:
On 15/07/16 07:50, Corneliu ZUZU wrote:
Create a common-side to establish, among others,
prototypes of
atomic functions called from common-code. Done to avoid introducing
inconsistencies between arch-side headers when we make
subtle
changes
implemented X86 functions in
common
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
Reviewed-by: Andrew Cooper <andrew.coop...@citrix.com>
---
Changed since v3:
* add wrapper over arm-side __atomic_add_unless() instead of removing the
leading underscores
---
xen/include/asm-
This wouldn't let me make a param of a function that used atomic_read() const.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
Reviewed-by: Andrew Cooper <andrew.coop...@citrix.com>
Reviewed-by: Stefano Stabellini <sstabell...@kernel.org>
---
Changed since v3:
---
x
(also updated README.LinuxPrimitives file).
Removed outdated comment ("NB. I've [...]").
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
Suggested-by: Andrew Cooper <andrew.coop...@citrix.com>
Reviewed-by: Andrew Cooper <andrew.coop...@citrix.com>
Reviewed-by:
Place atomic_inc_and_test() implementation after atomic_inc().
Also empty line fix.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
Reviewed-by: Andrew Cooper <andrew.coop...@citrix.com>
---
Changed since v3:
---
xen/include/asm-x86/atomic.h | 39 +++---
Move duplicate macros between asm-arm/arm32/atomic.h and asm-arm/arm64/atomic.h
to asm-arm/atomic.h. Adjust README.LinuxPrimitives in the process.
Also empty line fixes.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
Reviewed-by: Stefano Stabellini <sstabell...@kernel.org>
---
C
Corneliu ZUZU (7):
1/7: asm-arm/atomic.h: fix arm32|arm64 macros duplication
Reviewed-by: Stefano Stabellini <sstabell...@kernel.org>
Changed:
* update README.LinuxPrimitives file
2/7: asm-x86/atomic.h: minor: proper atomic_inc_and_test() placement
Reviewed-by:
On 7/14/2016 12:26 PM, Julien Grall wrote:
On 14/07/16 06:11, Corneliu ZUZU wrote:
On 7/13/2016 10:12 PM, Julien Grall wrote:
Hi Corneliu,
On 13/07/2016 15:18, Corneliu ZUZU wrote:
Move duplicate macros between asm-arm/arm32/atomic.h and
asm-arm/arm64/atomic.h
to asm-arm/atomic.h.
asm
On 7/14/2016 1:00 PM, Julien Grall wrote:
Hi Corneliu,
On 14/07/16 10:14, Corneliu ZUZU wrote:
The built-in leading underscores ('__') don't serve any purpose, so
rename
__atomic_add_unless() -> atomic_add_unless().
The leading underscores are from the Linux code. We decided to keep
th
On 7/14/2016 12:44 PM, Andrew Cooper wrote:
On 14/07/16 10:39, Corneliu ZUZU wrote:
On 7/14/2016 12:29 PM, Andrew Cooper wrote:
On 14/07/16 10:14, Corneliu ZUZU wrote:
- implement missing functions atomic_{sub,inc,dec}_return(), atomic_add_unless()
on X86 and also add prototypes for them
On 7/14/2016 12:29 PM, Andrew Cooper wrote:
On 14/07/16 10:14, Corneliu ZUZU wrote:
- implement missing functions atomic_{sub,inc,dec}_return(), atomic_add_unless()
on X86 and also add prototypes for them in common
- add missing macro atomic_xchg for X86
Signed-off-by: Corneliu ZUZU<
On 7/14/2016 12:26 PM, Andrew Cooper wrote:
On 14/07/16 10:13, Corneliu ZUZU wrote:
This wouldn't let me make a param of a function that used atomic_read() const.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
Reviewed-by: Andrew Cooper <andrew.coop...@citrix.com>
You
On 7/14/2016 12:26 PM, Andrew Cooper wrote:
On 14/07/16 10:13, Corneliu ZUZU wrote:
This wouldn't let me make a param of a function that used atomic_read() const.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
Reviewed-by: Andrew Cooper <andrew.coop...@citrix.com>
You
- implement missing functions atomic_{sub,inc,dec}_return(), atomic_add_unless()
on X86 and also add prototypes for them in common
- add missing macro atomic_xchg for X86
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/include/asm-x86/atomic.
The built-in leading underscores ('__') don't serve any purpose, so rename
__atomic_add_unless() -> atomic_add_unless().
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/include/asm-arm/arm32/atomic.h | 2 +-
xen/include/asm-arm/arm64/atomic.h | 2 +-
2 files changed, 2 i
Turn atomic_inc_return and atomic_dec_return atomic.h macros to inline
functions.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/include/asm-arm/atomic.h | 13 ++---
1 file changed, 10 insertions(+), 3 deletions(-)
diff --git a/xen/include/asm-arm/atomic.h b/xen/i
This wouldn't let me make a param of a function that used atomic_read() const.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
Reviewed-by: Andrew Cooper <andrew.coop...@citrix.com>
---
Changed since v2:
---
xen/include/asm-arm/atomic.h | 2 +-
xen/include/asm-x86/atomic.h
.
Removed outdated comment ("NB. I've [...]").
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
Suggested-by: Andrew Cooper <andrew.coop...@citrix.com>
Reviewed-by: Andrew Cooper <andrew.coop...@citrix.com>
---
Changed since v2:
* sample code for atomic_cmpxchg
Reorder macro definitions to match x86-side.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
Reviewed-by: Stefano Stabellini <sstabell...@kernel.org>
---
Changed since v2:
---
xen/include/asm-arm/atomic.h | 19 +--
1 file changed, 9 insertions(+), 10 deleti
Place atomic_inc_and_test() implementation after atomic_inc().
Also empty line fix.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
Changed since v2: arm empty line fixes moved from here
---
xen/include/asm-x86/atomic.h | 39 +++
1 file chang
Move duplicate macros between asm-arm/arm32/atomic.h and asm-arm/arm64/atomic.h
to asm-arm/atomic.h.
Also empty line fixes.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
Reviewed-by: Stefano Stabellini <sstabell...@kernel.org>
---
Changed since v2:
* also moved mistak
Corneliu ZUZU (8):
1/8: asm-arm/atomic.h: fix arm32|arm64 macros duplication
Reviewed-by: Stefano Stabellini <sstabell...@kernel.org>
Changed:
* also moved mistakenly omitted atomic_xchg()
* arm empty line fixes moved here
2/8: asm-x86/atomic.h: minor:
On 7/13/2016 10:12 PM, Julien Grall wrote:
Hi Corneliu,
On 13/07/2016 15:18, Corneliu ZUZU wrote:
Move duplicate macros between asm-arm/arm32/atomic.h and
asm-arm/arm64/atomic.h
to asm-arm/atomic.h.
asm-arm/arm*/atomic.h were a copy from Linux. I don't mind if we
diverge, however the file
On 7/13/2016 10:49 PM, Andrew Cooper wrote:
On 13/07/16 20:33, Corneliu ZUZU wrote:
On 7/13/2016 10:01 PM, Stefano Stabellini wrote:
On Wed, 13 Jul 2016, Corneliu ZUZU wrote:
Create a common-side to establish, among others,
prototypes of
atomic functions called from common-code. Done
Hi Julien,
On 7/13/2016 10:12 PM, Julien Grall wrote:
Hi Corneliu,
On 13/07/2016 15:18, Corneliu ZUZU wrote:
Move duplicate macros between asm-arm/arm32/atomic.h and
asm-arm/arm64/atomic.h
to asm-arm/atomic.h.
asm-arm/arm*/atomic.h were a copy from Linux. I don't mind if we
diverge
On 7/13/2016 10:01 PM, Stefano Stabellini wrote:
On Wed, 13 Jul 2016, Corneliu ZUZU wrote:
Create a common-side to establish, among others, prototypes of
atomic functions called from common-code. Done to avoid introducing
inconsistencies between arch-side headers when we make subtle
changes
.
Removed outdated comment ("NB. I've [...]").
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
Suggested-by: Andrew Cooper <andrew.coop...@citrix.com>
Reviewed-by: Andrew Cooper <andrew.coop...@citrix.com>
---
Changed since v1:
* removed comments that were duplic
This wouldn't let me make a param of a function that used atomic_read() const.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
Reviewed-by: Andrew Cooper <andrew.coop...@citrix.com>
---
Changed since v1:
---
xen/include/asm-arm/atomic.h | 2 +-
xen/include/asm-x86/atomic.h
Reorder macro definitions to match x86-side.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/include/asm-arm/atomic.h | 19 +--
1 file changed, 9 insertions(+), 10 deletions(-)
diff --git a/xen/include/asm-arm/atomic.h b/xen/include/asm-arm/atomic.h
index 1
Place atomic_inc_and_test() implementation after atomic_inc().
Also remove unneeded empty line and add a needed one.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/include/asm-arm/atomic.h | 1 +
xen/include/asm-x86/atomic.h | 39 +++
2
Move duplicate macros between asm-arm/arm32/atomic.h and asm-arm/arm64/atomic.h
to asm-arm/atomic.h.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/include/asm-arm/arm32/atomic.h | 11 ---
xen/include/asm-arm/arm64/atomic.h | 11 ---
xen/include/asm-arm/at
Corneliu ZUZU (5):
asm-arm/atomic.h: fix arm32|arm64 macros duplication
asm-x86/atomic.h: minor: proper atomic_inc_and_test() placement
asm-arm/atomic.h: reorder macros to match x86-side
asm/atomic.h: common prototyping (add xen/atomic.h)
fix: make atomic_read() param const
xen/include
On 7/13/2016 3:12 PM, Andrew Cooper wrote:
On 13/07/16 12:23, Corneliu ZUZU wrote:
+typedef struct { int counter; } atomic_t;
+
+#define ATOMIC_INIT(i) { (i) }
+
+/**
+ * atomic_read - read atomic variable
+ * @v: pointer of type atomic_t
+ *
+ * Atomically reads the value of @v.
+ */
+static
Hi Julien,
On 7/13/2016 3:14 PM, Julien Grall wrote:
Hi Corneliu,
On 13/07/2016 12:23, Corneliu ZUZU wrote:
Following Andrew Cooper's suggestion, create a common-side
to
establish, among others, prototypes of atomic functions called from
common-code.
Done to avoid introducing
On 7/13/2016 3:12 PM, Andrew Cooper wrote:
On 13/07/16 12:23, Corneliu ZUZU wrote:
Following Andrew Cooper's suggestion, create a common-side to
establish, among others, prototypes of atomic functions called from common-code.
Done to avoid introducing inconsistencies between arch-side
headers
This wouldn't let me make a param of a function that used atomic_read() const.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/include/asm-arm/atomic.h | 2 +-
xen/include/asm-x86/atomic.h | 2 +-
xen/include/xen/atomic.h | 2 +-
3 files changed, 3 insertions(+), 3 del
into inline functions in the process.
Also includes a minor adjustment asm-x86/atomic.h: reorder atomic_inc_and_test()
to follow after atomic_inc().
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/include/asm-arm/arm32/atomic.h | 11 ---
xen/include/asm-arm/arm64/atomic.h
Corneliu ZUZU (2):
asm/atomic.h: common prototyping (add xen/atomic.h)
fix: make atomic_read() param const
xen/include/asm-arm/arm32/atomic.h | 11 ---
xen/include/asm-arm/arm64/atomic.h | 11 ---
xen/include/asm-arm/atomic.h | 91 +++---
xen/include/asm-x86/atomic.h
On 7/9/2016 9:57 PM, Corneliu ZUZU wrote:
On 7/9/2016 9:26 PM, Tamas K Lengyel wrote:
diff --git a/xen/include/asm-x86/domain.h
b/xen/include/asm-x86/domain.h
index ae1dcb4..7663da2 100644
--- a/xen/include/asm-x86/domain.h
+++ b/xen/include/asm-x86/domain.h
@@ -9,6 +9,7 @@
#include
On 7/12/2016 3:49 PM, Andrew Cooper wrote:
On 12/07/16 11:38, Corneliu ZUZU wrote:
On 7/12/2016 1:22 PM, Andrew Cooper wrote:
On 12/07/16 11:11, Corneliu ZUZU wrote:
If you are really feeling up to it, having a common xen/atomic.h
with
typedef struct { int counter; } atomic_t;
#define
On 7/12/2016 1:22 PM, Andrew Cooper wrote:
On 12/07/16 11:11, Corneliu ZUZU wrote:
If you are really feeling up to it, having a common xen/atomic.h with
typedef struct { int counter; } atomic_t;
#define ATOMIC_INIT(i) { (i) }
and some prototypes such as:
static inline int atomic_read(const
On 7/12/2016 1:22 PM, Andrew Cooper wrote:
On 12/07/16 11:11, Corneliu ZUZU wrote:
If you are really feeling up to it, having a common xen/atomic.h with
typedef struct { int counter; } atomic_t;
#define ATOMIC_INIT(i) { (i) }
and some prototypes such as:
static inline int atomic_read(const
On 7/12/2016 12:42 PM, Andrew Cooper wrote:
On 12/07/16 06:11, Corneliu ZUZU wrote:
Hi Andrew,
On 7/11/2016 6:18 PM, Andrew Cooper wrote:
On 09/07/16 05:12, Corneliu ZUZU wrote:
This wouldn't let me make a param of a function that used
atomic_read() const.
Signed-off-by: Corneliu ZUZU <
On 7/12/2016 10:45 AM, Tian, Kevin wrote:
From: Corneliu ZUZU [mailto:cz...@bitdefender.com]
Sent: Monday, July 11, 2016 2:19 PM
+static inline
+void monitor_ctrlreg_adjust_traps(struct domain *d, unsigned int index)
+{
+/* For now, VMX only. */
+ASSERT(cpu_has_vmx);
+
+/* Other CRs
On 7/11/2016 7:43 PM, Tamas K Lengyel wrote:
On Sat, Jul 9, 2016 at 12:46 PM, Corneliu ZUZU <cz...@bitdefender.com> wrote:
On 7/9/2016 9:10 PM, Tamas K Lengyel wrote:
On Fri, Jul 8, 2016 at 10:13 PM, Corneliu ZUZU <cz...@bitdefender.com>
wrote:
Arch-specific vm-event func
Hi Andrew,
On 7/11/2016 6:18 PM, Andrew Cooper wrote:
On 09/07/16 05:12, Corneliu ZUZU wrote:
This wouldn't let me make a param of a function that used atomic_read() const.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
This is a good improvement, but you must make an ide
On 7/12/2016 12:27 AM, Tamas K Lengyel wrote:
On Mon, Jul 11, 2016 at 2:20 PM, Corneliu ZUZU <cz...@bitdefender.com> wrote:
On 7/11/2016 7:38 PM, Tamas K Lengyel wrote:
diff --git a/xen/include/asm-arm/monitor.h
b/xen/include/asm-arm/monitor.h
index 9a9734a..7ef30f1 100644
[snip]
On 7/11/2016 7:38 PM, Tamas K Lengyel wrote:
diff --git a/xen/include/asm-arm/monitor.h
b/xen/include/asm-arm/monitor.h
index 9a9734a..7ef30f1 100644
[snip]
I keep seeing '[snip]' lately but I don't know what it means.
Placeholder for "I've cut some of your patch content here to shorten
the
This was 1/8 in the old v3-series "x86/vm-event: Adjustments & fixes".
Transferring code-review from "Tian, Kevin <kevin.t...@intel.com>" (July
11, 2016).
On 7/9/2016 7:12 AM, Corneliu ZUZU wrote:
Minor optimization @ vmx_update_guest_cr: checks if v->arch.
This was 2/8 in the old v3-series "x86/vm-event: Adjustments & fixes".
Transferring code-review from "Tian, Kevin <kevin.t...@intel.com>" (July
11, 2016) and responding...
On 7/9/2016 7:15 AM, Corneliu ZUZU wrote:
For readability:
* Add function monito
On 7/11/2016 5:54 AM, Tian, Kevin wrote:
what's the difference between this series and earlier one?
[PATCH v3 0/8] x86/vm-event: Adjustments & fixes
looks you have some patches cross-posted (e.g. 1/16)...
Corneliu ZUZU (16):
x86/vmx_update_guest_cr: minor optimization
x86: fix:
On 7/9/2016 9:26 PM, Tamas K Lengyel wrote:
diff --git a/xen/include/asm-x86/domain.h b/xen/include/asm-x86/domain.h
index ae1dcb4..7663da2 100644
--- a/xen/include/asm-x86/domain.h
+++ b/xen/include/asm-x86/domain.h
@@ -9,6 +9,7 @@
#include
#include
#include
+#include
#include
On 7/9/2016 9:14 PM, Tamas K Lengyel wrote:
On Fri, Jul 8, 2016 at 10:14 PM, Corneliu ZUZU <cz...@bitdefender.com> wrote:
vm_event_register_write_resume is part of the monitor subsystem, relocate and
rename appropriately.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
On 7/9/2016 9:10 PM, Tamas K Lengyel wrote:
On Fri, Jul 8, 2016 at 10:13 PM, Corneliu ZUZU <cz...@bitdefender.com> wrote:
Arch-specific vm-event functions in x86/vm_event.h -e.g. vm_event_init_domain()-
don't have an 'arch_' prefix. Apply the same rule for monitor functions -
orig
On 7/9/2016 8:34 PM, Tamas K Lengyel wrote:
diff --git a/xen/arch/x86/monitor.c b/xen/arch/x86/monitor.c
index aeee435..4a29cad 100644
--- a/xen/arch/x86/monitor.c
+++ b/xen/arch/x86/monitor.c
@@ -62,6 +62,8 @@ int monitor_init_domain(struct domain *d)
return -ENOMEM;
}
+
On 7/9/2016 7:23 AM, Corneliu ZUZU wrote:
Enforce presence of a monitor vm-event subscriber when the toolstack user calls
xc_monitor_write_ctrlreg() (XEN_DOMCTL_MONITOR_EVENT_WRITE_CTRLREG domctl).
Without this change, "ASSERT(monitor_domain_initialised(v->domain));"
ck user calls
xc_monitor_write_ctrlreg(...) w/ enable = true, without first calling
xc_monitor_enable().
Also adjust returned error code for similar check from -EINVAL to more
descriptive -ENODEV (XEN_DOMCTL_MONITOR_OP_EMULATE_EACH_REP).
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen
up_domain() was called,
arch_vm_event.monitor will eventually be freed either after the CR-writes are
committed (monitor_ctrlreg_write_data()) or when the domain/vcpu is destroyed
* calls monitor_ctrlreg_write_data() even if the monitor subsystem is not
initialized
Signed-off-by: Corneliu
h and asm-x86/monitor.h.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/arch/x86/hvm/emulate.c | 8 +++---
xen/arch/x86/hvm/hvm.c | 31
xen/arch/x86/mm/p2m.c | 7 --
xen/arch/x86/mon
Introduce writes_pending field in monitor_write_data structure to slightly
optimize code @ monitor_write_data(): avoid having to check each bit when -all-
bits are zero (which is likely).
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/arch/x86/monitor.c | 3 +++
xen/i
Move cleanup of mem_access_emulate_each_rep to monitor_cleanup_domain() as the
field is part of the monitor subsystem's resources.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/arch/x86/monitor.c | 3 +++
xen/arch/x86/vm_event.c | 2 --
2 files changed, 3 insertions
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/arch/x86/hvm/emulate.c| 3 ++-
xen/arch/x86/hvm/hvm.c| 10 +-
xen/arch/x86/monitor.c| 3 +++
xen/include/asm-arm/monitor.h | 3 ++-
xen/include/asm-x86/monitor.h | 9 -
xen/include/xen/monito
Move vm-event cleanup sequence on vcpu destroyal to vm_event.h along with the
other vm-event functions.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/arch/x86/domain.c | 4 ++--
xen/arch/x86/vm_event.c| 5 +
xen/include/asm-x86/vm_event.h | 10 +++
d->monitor is a monitor subsystem resource, clean it up in the proper stub.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/include/asm-arm/monitor.h | 2 +-
xen/include/asm-arm/vm_event.h | 2 --
2 files changed, 1 insertion(+), 3 deletions(-)
diff --git a/xen/includ
relying on the
ved param in vm_event_init,cleanup_domain() to determine the target vm-event
subsystem when we already had vec->mode in vm_event_enable().
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/arch/x86/monitor.c | 5 +
xen/arch/x86/vm_event.c
From vm_event_init_domain(d), call vm_event_cleanup_domain(d) if per-vcpu
allocations failed -at some point- to cleanup partial allocations, if any were
done along the way.
Also minor adjustments: add unlikely in if's and add some comments.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.
gly.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/arch/x86/monitor.c | 10 ++
1 file changed, 6 insertions(+), 4 deletions(-)
diff --git a/xen/arch/x86/monitor.c b/xen/arch/x86/monitor.c
index 29188e5..8f41f21 100644
--- a/xen/arch/x86/monitor.c
+++ b/xen/arch/x86/
-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/arch/x86/hvm/hvm.c| 46 ++
xen/arch/x86/hvm/vmx/vmx.c| 59 ++---
xen/arch/x86/monitor.c| 68 +++
xen/include/asm-x86/hvm/vmx
vm_event_register_write_resume is part of the monitor subsystem, relocate and
rename appropriately.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/arch/x86/monitor.c | 38 ++
xen/arch/x86/vm_event.c
but those 2 aforementioned).
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/arch/x86/monitor.c| 4 ++--
xen/common/monitor.c | 4 ++--
xen/common/vm_event.c | 4 ++--
xen/include/asm-arm/monitor.h | 8 +++-
xen/include/asm-x86/monitor.h | 12 ++---
This wouldn't let me make a param of a function that used atomic_read() const.
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/include/asm-x86/atomic.h | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/xen/include/asm-x86/atomic.h b/xen/include/asm-x86/at
Minor optimization @ vmx_update_guest_cr: checks if v->arch.hvm_vmx.exec_control
was modified before actually calling vmx_update_cpu_exec_control(v).
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
xen/arch/x86/hvm/vmx/vmx.c | 5 -
1 file changed, 4 insertions(+), 1
Adjustments & fixes to the vm-events code, mostly monitor-related.
As I hadn't got the time/proper context to test these changes on a real machine,
for now please consider them only for review, I'll let you know when and how
actual testing went.
Thanks,
Corneliu ZUZU (16):
On 7/8/2016 6:50 PM, Tamas K Lengyel wrote:
diff --git a/xen/arch/x86/monitor.c b/xen/arch/x86/monitor.c
index 205df41..42f31bf 100644
--- a/xen/arch/x86/monitor.c
+++ b/xen/arch/x86/monitor.c
[snip]
So with this code-portion appropriately being relocated here I think
we should also rename..
On 7/6/2016 6:49 PM, Corneliu ZUZU wrote:
Minor optimization @ vmx_update_guest_cr: checks if v->arch.hvm_vmx.exec_control
was modified before actually calling vmx_update_cpu_exec_control(v).
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
Changed since v2:
---
xen/arch/x8
On 7/8/2016 2:53 PM, Jan Beulich wrote:
On 08.07.16 at 13:33, wrote:
On 7/8/2016 1:37 PM, Jan Beulich wrote:
On 08.07.16 at 12:22, wrote:
On 7/8/2016 10:21 AM, Jan Beulich wrote:
On 06.07.16 at 17:50, wrote:
+/*
On 7/8/2016 2:48 PM, Jan Beulich wrote:
On 08.07.16 at 13:39, <cz...@bitdefender.com> wrote:
On 7/6/2016 6:49 PM, Corneliu ZUZU wrote:
Minor optimization @ vmx_update_guest_cr: checks if
v->arch.hvm_vmx.exec_control
was modified before actually calling vmx_update_cpu_exec_control(v)
On 7/6/2016 6:49 PM, Corneliu ZUZU wrote:
Minor optimization @ vmx_update_guest_cr: checks if v->arch.hvm_vmx.exec_control
was modified before actually calling vmx_update_cpu_exec_control(v).
Signed-off-by: Corneliu ZUZU <cz...@bitdefender.com>
---
Changed since v2:
---
xen/arch/x8
On 7/8/2016 1:37 PM, Jan Beulich wrote:
On 08.07.16 at 12:22, wrote:
On 7/8/2016 10:21 AM, Jan Beulich wrote:
On 06.07.16 at 17:50, wrote:
The title of this patch keeps confusing me - which code motion is
being relocated here?
As the commit
On 7/8/2016 10:56 AM, Jan Beulich wrote:
On 06.07.16 at 17:55, wrote:
--- a/xen/arch/x86/hvm/hvm.c
+++ b/xen/arch/x86/hvm/hvm.c
@@ -31,7 +31,6 @@
#include
#include
#include
-#include
#include
#include
#include
@@ -63,11 +62,9 @@
#include
#include
On 7/8/2016 10:35 AM, Jan Beulich wrote:
On 06.07.16 at 17:51, wrote:
@@ -492,8 +493,12 @@ int vcpu_initialise(struct vcpu *v)
void vcpu_destroy(struct vcpu *v)
{
-xfree(v->arch.vm_event);
-v->arch.vm_event = NULL;
+if ( unlikely(v->arch.vm_event) )
+
On 7/8/2016 10:21 AM, Jan Beulich wrote:
On 06.07.16 at 17:50, wrote:
The title of this patch keeps confusing me - which code motion is
being relocated here?
As the commit message clearly states, the code motions that are being
relocated are:
1) handling of
On 7/7/2016 11:27 AM, Jan Beulich wrote:
On 06.07.16 at 17:54, wrote:
--- a/xen/arch/x86/hvm/hvm.c
+++ b/xen/arch/x86/hvm/hvm.c
@@ -475,7 +475,7 @@ void hvm_do_resume(struct vcpu *v)
if ( unlikely(v->arch.vm_event) )
{
-if (
1 - 100 of 314 matches
Mail list logo