[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [Xen-devel] [PATCH v10 00/20] x86/PMU: Xen PMU PV(H) support



>>> On 04.09.14 at 05:41, <boris.ostrovsky@xxxxxxxxxx> wrote:
> Here is the tenth version of PV(H) PMU patches.
> 
> Changes in v10:
> 
> * Swapped address and name fields of xenpf_symdata (to make it smaller on 
> 32-bit)
> * Dropped vmx_rm_guest_msr() as it requires refcountig which makes code more 
> complicated.
> * Cleaned up vlapic_reg_write()
> * Call vpmu_destroy() for both HVM and PVH VCPUs
> * Verify that (xen_pmu_data+PMU register bank) fit into a page
> * Return error codes from arch-specific VPMU init code
> * Moved VPMU-related context switch logic into inlines
> * vpmu_force_context_switch() changes:
>   o Avoid greater than page-sized allocations
>   o Prevent another VCPU from starting VPMU sync while the first sync is in 
> progress
> * Avoid stack leak in do_xenpmu_op()
> * Checked validity of Intel VPMU MSR values before they are committed
> * Fixed MSR handling in traps.c (avoid potential accesses to Intel MSRs on 
> AMD)
> * Fixed VCPU selection in interrupt handler for 32-bit dom0 (sampled => 
> sampling)
> * Clarified commit messages (patches 2, 13, 18) 
> * Various cleanups

So I went though this again, and only very few issues seem to be left.
Assuming they will get addressed, the series could imo go in provided
all of its constituents (perhaps apart from the last one) have an
up-to-date Reviewed-by tag from someone deemed knowledgeable in
the area (I'm certainly not).

Jan


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel


 


Rackspace

Lists.xenproject.org is hosted with RackSpace, monitoring our
servers 24x7x365 and backed by RackSpace's Fanatical Support®.