Re: [PATCH 2/2] x86/vm_event: Allow returning i-cache for emulation

From: Razvan Cojocaru <rcojocaru@bitdefender.com>
To: Tamas K Lengyel <tamas.lengyel@zentific.com>
Cc: Kevin Tian <kevin.tian@intel.com>,
	Stefano Stabellini <sstabellini@kernel.org>,
	Jan Beulich <jbeulich@suse.com>,
	George Dunlap <george.dunlap@eu.citrix.com>,
	Andrew Cooper <andrew.cooper3@citrix.com>,
	Julien Grall <julien.grall@arm.com>,
	Paul Durrant <paul.durrant@citrix.com>,
	Jun Nakajima <jun.nakajima@intel.com>,
	"xen-devel@lists.xenproject.org" <xen-devel@lists.xenproject.org>
Subject: Re: [PATCH 2/2] x86/vm_event: Allow returning i-cache for emulation
Date: Fri, 16 Sep 2016 19:09:55 +0300	[thread overview]
Message-ID: <8af58cfa-59e6-ade5-44ec-826894e35b8a@bitdefender.com> (raw)
In-Reply-To: <CAErYnshRAKuDAC8QGGcVdQ0JX7qauP=0O4VKCp-83P-M3ar1yQ@mail.gmail.com>

On 09/16/16 18:37, Tamas K Lengyel wrote:
>>>>> Since this breaks compilation of existing clients, I think we should
>>>>> >>>> increment some macro to alow for compile-time switching (maybe
>>>>> >>>> VM_EVENT_INTERFACE_VERSION?).
>>>> >>>
>>>> >>> I'm not sure I follow - this is a Xen internal-only enumeration. What
>>>> >>> kind-of clients are you referring to?
>>> >>
>>> >> Nevermind, I thought these changes propagate to the toolstack headers.
>>> >> Sorry for the confusion.
>> >
>> > On second thought (and a night's rest), the problem is real. I've made
>> > it a point to test the patches today before re-reviewing them, and this
>> > happened:
>> >
>> > bdvmixeneventmanager.cpp:359:46: error: ‘union vm_event_st::<anonymous>’
>> > has no member named ‘emul_read_data’
>> >       uint32_t rspDataSize = sizeof( rsp.data.emul_read_data.data );
>> >                                               ^
>> > bdvmixeneventmanager.cpp:386:44: error: ‘union vm_event_st::<anonymous>’
>> > has no member named ‘emul_read_data’
>> >                            action, rsp.data.emul_read_data.data,
>> > rspDataSize,
>> >                                             ^
>> > bdvmixeneventmanager.cpp:389:16: error: ‘union vm_event_st::<anonymous>’
>> > has no member named ‘emul_read_data’
>> >        rsp.data.emul_read_data.size = rspDataSize;
>> >                 ^
>> > make: *** [bdvmixeneventmanager.o] Error 1
>> >
>> > So, as you can see, existing clients really don't compile without
>> > modification.
>> >
> Hi Razvan,
> yes, for Xen 4.8 we already bumped the VM_EVENT_INTERFACE_VERSION, so
> any client building on it need to adapt accordingly.

Fair enough. That's all I was asking for / about: that we should make
sure that VM_EVENT_INTERFACE_VERSION reflects these changes. If it's
already been incremented for this pending release, that's perfect.

Thanks,
Razvan

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