Re: [libvirt] [PATCH 4/5] qemu: Add -mem-path even with numa
On Thu, Oct 01, 2015 at 07:01:17PM -0400, John Ferlan wrote:
On 10/01/2015 08:10 AM, Martin Kletzander wrote:
> Of course this will be used only in case we don't need the
> memory-backend-file backend, so it should not fire until later.
>
So if I'm reading things right... when building the numa string, if it's
determine that "-object" and ",memdev=ram-node%zu" will be required,
then it's also required to have "-mem-alloc -mem-path ..."
(furrowing through the qemu docs was more difficult than I thought)
This was harder for me to grasp as well, I'll try to explain. If you
use -object memory-backend-file, that object itself has a path where
to use hugepages from. If you instead use '-numa ...,mem=SIZE', then
there is no way to say where from can this particular node take the
hugepages, but you can use '-mem-prealloc -mem-path
/path/to/hugetlbfs/mount' to specify it for the whole domain. So you
need '-mem-path' if hugepages backing is requested and there is no
need for -object memory-backend-file'.
I hope that's understandable, let me know if I should try explaining
it more clearly.
> Signed-off-by: Martin Kletzander <mkletzan(a)redhat.com>
> ---
> src/qemu/qemu_command.c | 4 ++++
> 1 file changed, 4 insertions(+)
>
Perhaps "outside" the domain of these changes; however, I note that the
call to qemuBuildNumaArgStr is from qemuBuildCommandLine as follows:
if (virDomainNumaGetNodeCount(def->numa)) {
if (qemuBuildNumaArgStr(cfg, def, cmd, qemuCaps, nodeset) < 0)
one of the first things done in qemuBuildNumaArgStr:
size_t ncells = virDomainNumaGetNodeCount(def->numa);
oh and of course the pesky
const long system_page_size = virGetSystemPageSizeKB();
I mentioned previously
BTW: Not noted by Coverity, but interestingly later in the code there's:
if (ncells) {
/* Fortunately, we allow only guest NUMA nodes to be continuous
* starting from zero. */
pos = ncells - 1;
}
Which if I'm reading things correct - cannot happen.
I haven't noticed this, but the way I'm reading it is that
qemuBuildNumaArgStr() is called only if there are some numa nodes, so
this will happen *every* time qemuBuildNumaArgStr().
Just a note - you don't have to change/fix this, but well if you
want to...
ACK to what's here as it seems reasonable - perhaps a better reason in
the commit message would help
I'll update the commit message, reading it after myself now I don't
understand it either :)
John
> diff --git a/src/qemu/qemu_command.c b/src/qemu/qemu_command.c
> index 17e5cfd71702..321a5e931350 100644
> --- a/src/qemu/qemu_command.c
> +++ b/src/qemu/qemu_command.c
> @@ -8132,6 +8132,10 @@ qemuBuildNumaArgStr(virQEMUDriverConfigPtr cfg,
> }
> }
>
> + if (!needBackend &&
> + qemuBuildMemPathStr(cfg, def, qemuCaps, cmd) < 0)
> + goto cleanup;
> +
> for (i = 0; i < ncells; i++) {
> VIR_FREE(cpumask);
> if (!(cpumask = virBitmapFormat(virDomainNumaGetNodeCpumask(def->numa,
i))))
>
Attachments:
- signature.asc (application/pgp-signature — 819 bytes)