[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
Re: [PATCH] xen: Drop xenoprofile support
On 1/6/26 9:53 AM, Jan Beulich wrote:
On 06.01.2026 09:43, Oleksii Kurochko wrote:
On 1/5/26 8:57 PM, Andrew Cooper wrote:
The most recent xenoprof change was 300ef0cb4fde ("x86: Add Xenoprofile
support for AMD Family16h") in 2013, despite there being 42 changes worth of
other cleanup/rearranging since then.
Oprofile themselves dropped Xen support in commit 0c142c3a096d ("Remove
opcontrol and the GUI and processor models dependent on it") in 2014, as part
of releasing version 1.0 and switching over to using operf based on the Linux
perf_event subsystem. Linux's version of this patch was merged in commit
24880bef417f ("Merge tag 'oprofile-removal-5.12'") in 2021.
Drop xenoprof and all supporting infrastructure, including the hypercall, the
XSM hook and flask vectors which lose their only caller, and even shrinks
struct domain by one pointer which wasn't properly excluded in
!CONFIG_XENOPROF builds.
Retain the public xenoprof.h header as it is ABI, but note that the
functionality is removed.
Signed-off-by: Andrew Cooper <andrew.cooper3@xxxxxxxxxx>
---
CC: Anthony PERARD <anthony.perard@xxxxxxxxxx>
CC: Michal Orzel <michal.orzel@xxxxxxx>
CC: Jan Beulich <jbeulich@xxxxxxxx>
CC: Julien Grall <julien@xxxxxxx>
CC: Roger Pau Monné <roger.pau@xxxxxxxxxx>
CC: Stefano Stabellini <sstabellini@xxxxxxxxxx>
CC: Oleksii Kurochko <oleksii.kurochko@xxxxxxxxx>
CC: Daniel P. Smith <dpsmith@xxxxxxxxxxxxxxxxxxxx>
Despite appearing to be architecture neutral, the internals of Xenoprof were
entirely x86-specific. Another curiosity is that only the VMX MSR hooks
called passive_domain_do_{rd,wr}msr(), and I can't see how this was correct
for SVM.
The real reason for finally getting around to this is the number of MISRA
violations reported by the eclair-x86_64-allcode job that I don't feel like
fixing.
---
CHANGELOG.md | 3 +
docs/misc/xen-command-line.pandoc | 6 -
tools/flask/policy/modules/dom0.te | 2 -
xen/arch/x86/Makefile | 1 -
xen/arch/x86/cpu/vpmu_amd.c | 7 -
xen/arch/x86/cpu/vpmu_intel.c | 6 -
xen/arch/x86/hvm/svm/entry.S | 1 -
xen/arch/x86/hvm/svm/svm.c | 2 -
xen/arch/x86/hvm/vmx/vmx.c | 9 -
xen/arch/x86/include/asm/xenoprof.h | 95 ---
xen/arch/x86/oprofile/Makefile | 6 -
xen/arch/x86/oprofile/backtrace.c | 145 ----
xen/arch/x86/oprofile/nmi_int.c | 485 ------------
xen/arch/x86/oprofile/op_counter.h | 41 -
xen/arch/x86/oprofile/op_model_athlon.c | 547 -------------
xen/arch/x86/oprofile/op_model_p4.c | 721 -----------------
xen/arch/x86/oprofile/op_model_ppro.c | 348 ---------
xen/arch/x86/oprofile/op_x86_model.h | 58 --
xen/arch/x86/oprofile/xenoprof.c | 106 ---
xen/arch/x86/traps.c | 4 -
xen/common/Kconfig | 11 -
xen/common/Makefile | 1 -
xen/common/compat/xenoprof.c | 42 -
xen/common/domain.c | 6 -
xen/common/xenoprof.c | 977 ------------------------
xen/include/Makefile | 1 -
xen/include/hypercall-defs.c | 6 -
xen/include/public/xen.h | 2 +-
xen/include/public/xenoprof.h | 2 +-
xen/include/xen/sched.h | 3 -
xen/include/xen/xenoprof.h | 49 --
xen/include/xsm/dummy.h | 7 -
xen/include/xsm/xsm.h | 7 -
xen/xsm/dummy.c | 2 -
xen/xsm/flask/hooks.c | 35 -
xen/xsm/flask/policy/access_vectors | 4 -
36 files changed, 5 insertions(+), 3743 deletions(-)
delete mode 100644 xen/arch/x86/include/asm/xenoprof.h
delete mode 100644 xen/arch/x86/oprofile/Makefile
delete mode 100644 xen/arch/x86/oprofile/backtrace.c
delete mode 100644 xen/arch/x86/oprofile/nmi_int.c
delete mode 100644 xen/arch/x86/oprofile/op_counter.h
delete mode 100644 xen/arch/x86/oprofile/op_model_athlon.c
delete mode 100644 xen/arch/x86/oprofile/op_model_p4.c
delete mode 100644 xen/arch/x86/oprofile/op_model_ppro.c
delete mode 100644 xen/arch/x86/oprofile/op_x86_model.h
delete mode 100644 xen/arch/x86/oprofile/xenoprof.c
delete mode 100644 xen/common/compat/xenoprof.c
delete mode 100644 xen/common/xenoprof.c
delete mode 100644 xen/include/xen/xenoprof.h
diff --git a/CHANGELOG.md b/CHANGELOG.md
index 3aaf5986231c..1663f6878ef2 100644
--- a/CHANGELOG.md
+++ b/CHANGELOG.md
@@ -15,6 +15,9 @@ The format is based on [Keep a
Changelog](https://keepachangelog.com/en/1.0.0/)
- The cpuid_mask_* command line options for legacy AMD CPUs. These were
deprecated in Xen 4.7 and noted not to work correctly with AMD CPUs from
2011 onwards.
+ - Xenoprofile support. Oprofile themselves removed support for Xen in 2014
+ prior to the version 1.0 release, and there has been no development since
+ before then in Xen.
LGTM:
Acked-by: Oleksii Kurochko <oleksii.kurochko@xxxxxxxxx> # CHANGELOG.md
Nit: It is necessary to drop the extra space before " Oprofile themselves...".
Why would that be? See the other bullet point in context, which also uses a
two blanks after the inner full stop. This is deliberate.
I just missed that a similar case was mentioned above. If this is deliberate,
then I’m fine with it. It’s just not obvious (at least, for me) where it’s
acceptable and where it isn’t to have extra spaces.
~ Oleksii
|
