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

Re: [Xen-devel] [PATCH] Fix cpumap setting before passing to XEN



On 2016/4/12 11:35, Zhenzhong Duan wrote:
On 2016/4/11 19:27, Wei Liu wrote:
On Mon, Apr 11, 2016 at 09:42:57AM +0800, Zhenzhong Duan wrote:
It's tool's duty to pass a correct cpumap to XEN. On a host with less than
64
CPUS, it just shows below error.

[root@localhost /]# xm vcpu-pin 3 all all
Error: Cannot pin vcpu: 0 to cpu: [0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49, 50,
51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63] - (22, 'Invalid
argument')

This error clearly shows that the upper layer is not passing down the
right cpumap. So this example along won't convince me to that this is a
bug in python binding. It's free to reject input that it deems wrong.

I guess the question is more or less what should it do in such
situation. So ...

The fix make it same as in xl code.

Is there reference to this? I guess the commit message  or mail thread
will shed more light on the expected behaviour.
Calling path:
main_vcpupin->cpurange_parse->update_cpumap_range->libxl_bitmap_set

void libxl_bitmap_set(libxl_bitmap *bitmap, int bit)
{
    if (bit >= bitmap->size * 8)
        return;
    bitmap->map[bit / 8] |= 1 << (bit & 7);
}

I referenced above code. It just ignore the cpu number beyond the max CPU count rather than raise an error.
Any further comments?

thanks
zduan

_______________________________________________
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®.