> On Fri, 2010-09-17 at 07:10 +0100, Juergen Gross wrote:
>
> attached patch adds support of cpu pools in xl/libxl.
> Ian, I didn't split up this patch, because the updated libxl_cpumask
> structure requires changes to xl in any case.
That's fine.
> BTW: I realized that libxlu_cfg_l.c seems to be in the repository.
> Isn't this file generated by make? I think it should be removed there.
In an ideal world, yes. I think this was a workaround for buggy versions
of flex observed in the field or something.
> Signed-off-by: juergen.gross@xxxxxxxxxxxxxx
>
> Support of cpu pools in xl:
> library functions
> xl pool-create
> xl pool-list
> xl pool-destroy
> xl pool-cpu-add
> xl pool-cpu-remove
> xl pool-migrate
>
> diff -r d978675f3d53 tools/libxl/libxl.c
> --- a/tools/libxl/libxl.c Thu Sep 16 18:29:26 2010 +0100
> +++ b/tools/libxl/libxl.c Fri Sep 17 07:42:30 2010 +0200
> @@ -609,9 +609,17 @@ libxl_poolinfo * libxl_list_pool(libxl_c
> libxl_poolinfo * libxl_list_pool(libxl_ctx *ctx, int *nb_pool)
> {
> libxl_poolinfo *ptr;
> - int i, ret;
> - xc_cpupoolinfo_t info[256];
> - int size = 256;
> + int i, m;
> + xc_cpupoolinfo_t *info;
> + int size;
> + uint32_t poolid;
> + libxl_physinfo physinfo;
> +
> + if (libxl_get_physinfo(ctx, &physinfo) != 0) {
> + LIBXL__LOG_ERRNO(ctx, LIBXL__LOG_ERROR, "getting phys info");
> + return NULL;
> + }
> + size = physinfo.max_cpu_id + 32;
Where does the number 32 come from?
> ptr = calloc(size, sizeof(libxl_poolinfo));
> if (!ptr) {
> @@ -619,16 +627,23 @@ libxl_poolinfo * libxl_list_pool(libxl_c
> return NULL;
> }
>
> - ret = xc_cpupool_getinfo(ctx->xch, 0, 256, info);
> - if (ret<0) {
> - LIBXL__LOG_ERRNO(ctx, LIBXL__LOG_ERROR, "getting cpupool info");
> - return NULL;
> + poolid = 0;
> + for (i = 0; i < size; i++) {
> + info = xc_cpupool_getinfo(ctx->xch, poolid);
> + if (info == NULL)
> + break;
> + ptr[i].poolid = info->cpupool_id;
> + ptr[i].sched_id = info->sched_id;
> + ptr[i].n_dom = info->n_dom;
> + if (libxl_cpumap_alloc(&ptr[i].cpumap, physinfo.max_cpu_id + 1))
> + break;
> + for (m = 0; m < ptr[i].cpumap.size / sizeof(*ptr[i].cpumap.map); m++)
> + ptr[i].cpumap.map[m] = info->cpumap[m];
I guess if "physinfo.max_cpu_id + 1" does not correspond to
info->cpumap_size then that is a serious error but if it does happen
then you may run over the end of either ptr[i].cpumap.map or
info->cpumap.
Ian.
_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxxxxxxxx
http://lists.xensource.com/xen-devel
|
Home