[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] Re: [Xen-devel] [PATCH 04/18] xen: Preserve reserved grant entries when switching versions
On Thu, 2012-01-12 at 09:49 +0000, Ian Campbell wrote: > On Thu, 2012-01-12 at 08:53 +0000, Jan Beulich wrote: > > >>> On 11.01.12 at 18:21, Daniel De Graaf <dgdegra@xxxxxxxxxxxxx> wrote: > > > --- a/xen/common/grant_table.c > > > +++ b/xen/common/grant_table.c > > > @@ -2106,6 +2106,7 @@ > > > gnttab_set_version(XEN_GUEST_HANDLE(gnttab_set_version_t uop)) > > > struct domain *d = current->domain; > > > struct grant_table *gt = d->grant_table; > > > struct active_grant_entry *act; > > > + grant_entry_v1_t reserved_entries[GNTTAB_NR_RESERVED_ENTRIES]; > > > > How much stack space does this consume? (Or in other words, where > > does GNTTAB_NR_RESERVED_ENTRIES get defined? It's not in this > > patch, I can't find it in earlier patches, and it also doesn't exist in the > > current staging tree.) > > I could have sworn I'd seen this somewhere but I can't for the life of > me find it now :-/ Found it in patch #6. +/* External tools reserve first few grant table entries. */ +#define GNTTAB_NR_RESERVED_ENTRIES 8 +#define GNTTAB_RESERVED_CONSOLE 0 +#define GNTTAB_RESERVED_XENSTORE 1 +/* (the next 6 are reserved for future use) */ These patches need to be reordered to make sure each step builds. Ian. > > One place I did find it is in the Linux gnttab driver, but that's not > canonical. > > Ian. > > > > > Jan > > > > > long res; > > > int i; > > > > > > @@ -2126,7 +2127,7 @@ > > > gnttab_set_version(XEN_GUEST_HANDLE(gnttab_set_version_t uop)) > > > /* (You need to change the version number for e.g. kexec.) */ > > > if ( gt->gt_version != 0 ) > > > { > > > - for ( i = 0; i < nr_grant_entries(gt); i++ ) > > > + for ( i = GNTTAB_NR_RESERVED_ENTRIES; i < nr_grant_entries(gt); > > > i++ > > > ) > > > { > > > act = &active_entry(gt, i); > > > if ( act->pin != 0 ) > > > @@ -2151,15 +2152,38 @@ > > > gnttab_set_version(XEN_GUEST_HANDLE(gnttab_set_version_t uop)) > > > goto out_unlock; > > > } > > > > > > + /* Preserve the first 8 entries (toolstack reserved grants) */ > > > + if (gt->gt_version == 1) { > > > + memcpy(reserved_entries, gt->shared_v1[0], > > > sizeof(reserved_entries)); > > > + } else if (gt->gt_version == 2) { > > > + for ( i = 0; i < GNTTAB_NR_RESERVED_ENTRIES && i < > > > nr_grant_entries(gt); i++ ) > > > + { > > > + reserved_entries[i].flags = shared_entry_v2(gt, i).hdr.flags; > > > + reserved_entries[i].domid = shared_entry_v2(gt, i).hdr.domid; > > > + reserved_entries[i].frame = shared_entry_v2(gt, > > > i).full_page.frame; > > > + reserved_entries[i].flags |= status_entry(gt, i); > > > + } > > > + } > > > + > > > if ( op.version < 2 && gt->gt_version == 2 ) > > > gnttab_unpopulate_status_frames(d, gt); > > > > > > - if ( op.version != gt->gt_version ) > > > - { > > > - /* Make sure there's no crud left over in the table from the > > > - old version. */ > > > - for ( i = 0; i < nr_grant_frames(gt); i++ ) > > > - memset(gt->shared_raw[i], 0, PAGE_SIZE); > > > + /* Make sure there's no crud left over in the table from the > > > + old version. */ > > > + for ( i = 0; i < nr_grant_frames(gt); i++ ) > > > + memset(gt->shared_raw[i], 0, PAGE_SIZE); > > > + > > > + /* Restore the first 8 entries (toolstack reserved grants) */ > > > + if (gt->gt_version != 0 && op.version == 1) { > > > + memcpy(gt->shared_v1[0], reserved_entries, > > > sizeof(reserved_entries)); > > > + } else if (gt->gt_version != 0 && op.version == 2) { > > > + for ( i = 0; i < GNTTAB_NR_RESERVED_ENTRIES; i++ ) > > > + { > > > + status_entry(gt, i) = reserved_entries[i].flags & > > > (GTF_reading|GTF_writing); > > > + shared_entry_v2(gt, i).hdr.flags = reserved_entries[i].flags > > > & > > > ~(GTF_reading|GTF_writing); > > > + shared_entry_v2(gt, i).hdr.domid = reserved_entries[i].domid; > > > + shared_entry_v2(gt, i).full_page.frame = > > > reserved_entries[i].frame; > > > + } > > > } > > > > > > gt->gt_version = op.version; > > > -- > > > 1.7.7.5 > > > > > > > > > _______________________________________________ > > > Xen-devel mailing list > > > Xen-devel@xxxxxxxxxxxxxxxxxxx > > > http://lists.xensource.com/xen-devel > > > > > > > > > > _______________________________________________ > > Xen-devel mailing list > > Xen-devel@xxxxxxxxxxxxxxxxxxx > > http://lists.xensource.com/xen-devel > > > > _______________________________________________ > Xen-devel mailing list > Xen-devel@xxxxxxxxxxxxxxxxxxx > http://lists.xensource.com/xen-devel _______________________________________________ Xen-devel mailing list Xen-devel@xxxxxxxxxxxxxxxxxxx http://lists.xensource.com/xen-devel
|
Lists.xenproject.org is hosted with RackSpace, monitoring our |