| Message ID | 20210601161118.18986-6-olaf@aepfle.de (mailing list archive) |
|---|---|
| State | Superseded |
| Headers | show |
| Series | leftover from 2020 | expand |
On 01.06.21 18:10, Olaf Hering wrote: > Users of xc_get_pfn_type_batch may want to sanity check the data > returned by Xen. Add a simple helper for this purpose. > > Signed-off-by: Olaf Hering <olaf@aepfle.de> > --- > tools/libs/ctrl/xc_private.h | 33 +++++++++++++++++++++++++++++++++ > 1 file changed, 33 insertions(+) > > diff --git a/tools/libs/ctrl/xc_private.h b/tools/libs/ctrl/xc_private.h > index 5d2c7274fb..afb08aafe1 100644 > --- a/tools/libs/ctrl/xc_private.h > +++ b/tools/libs/ctrl/xc_private.h > @@ -421,6 +421,39 @@ void *xc_map_foreign_ranges(xc_interface *xch, uint32_t dom, > int xc_get_pfn_type_batch(xc_interface *xch, uint32_t dom, > unsigned int num, xen_pfn_t *); > > +/* Sanitiy check for types returned by Xen */ > +static inline bool xc_is_known_page_type(xen_pfn_t type) > +{ > + bool ret; > + > + switch (type) I think you should not imply the planned use case here. It would be better to use "switch (type & XEN_DOMCTL_PFINFO_LTAB_MASK)". I'm on the edge regarding putting the new function into xc_private.h. In the end your use case is _not_ to call the new function from libxenctrl. Juergen
Am Wed, 2 Jun 2021 08:51:45 +0200 schrieb Juergen Gross <jgross@suse.com>: > I think you should not imply the planned use case here. It would be > better to use "switch (type & XEN_DOMCTL_PFINFO_LTAB_MASK)". > > I'm on the edge regarding putting the new function into xc_private.h. > In the end your use case is _not_ to call the new function from > libxenctrl. I'm not sure what that means. One or the other has to be rebased to the new state. Olaf
On 02.06.21 13:10, Olaf Hering wrote: > Am Wed, 2 Jun 2021 08:51:45 +0200 > schrieb Juergen Gross <jgross@suse.com>: > >> I think you should not imply the planned use case here. It would be >> better to use "switch (type & XEN_DOMCTL_PFINFO_LTAB_MASK)". >> >> I'm on the edge regarding putting the new function into xc_private.h. >> In the end your use case is _not_ to call the new function from >> libxenctrl. > > > I'm not sure what that means. The name xc_private.h indicates that this header file is supposed to be used only inside of libxenctrl. I know that this isn't true today, but I don't think new misuses should be added to this file. > One or the other has to be rebased to the new state. You can add the same functions to some libxensaverestore prvate header instead. Juergen
diff --git a/tools/libs/ctrl/xc_private.h b/tools/libs/ctrl/xc_private.h index 5d2c7274fb..afb08aafe1 100644 --- a/tools/libs/ctrl/xc_private.h +++ b/tools/libs/ctrl/xc_private.h @@ -421,6 +421,39 @@ void *xc_map_foreign_ranges(xc_interface *xch, uint32_t dom, int xc_get_pfn_type_batch(xc_interface *xch, uint32_t dom, unsigned int num, xen_pfn_t *); +/* Sanitiy check for types returned by Xen */ +static inline bool xc_is_known_page_type(xen_pfn_t type) +{ + bool ret; + + switch (type) + { + case XEN_DOMCTL_PFINFO_NOTAB: + + case XEN_DOMCTL_PFINFO_L1TAB: + case XEN_DOMCTL_PFINFO_L1TAB | XEN_DOMCTL_PFINFO_LPINTAB: + + case XEN_DOMCTL_PFINFO_L2TAB: + case XEN_DOMCTL_PFINFO_L2TAB | XEN_DOMCTL_PFINFO_LPINTAB: + + case XEN_DOMCTL_PFINFO_L3TAB: + case XEN_DOMCTL_PFINFO_L3TAB | XEN_DOMCTL_PFINFO_LPINTAB: + + case XEN_DOMCTL_PFINFO_L4TAB: + case XEN_DOMCTL_PFINFO_L4TAB | XEN_DOMCTL_PFINFO_LPINTAB: + + case XEN_DOMCTL_PFINFO_XTAB: + case XEN_DOMCTL_PFINFO_XALLOC: + case XEN_DOMCTL_PFINFO_BROKEN: + ret = true; + break; + default: + ret = false; + break; + } + return ret; +} + void bitmap_64_to_byte(uint8_t *bp, const uint64_t *lp, int nbits); void bitmap_byte_to_64(uint64_t *lp, const uint8_t *bp, int nbits);
Users of xc_get_pfn_type_batch may want to sanity check the data returned by Xen. Add a simple helper for this purpose. Signed-off-by: Olaf Hering <olaf@aepfle.de> --- tools/libs/ctrl/xc_private.h | 33 +++++++++++++++++++++++++++++++++ 1 file changed, 33 insertions(+)