[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index] Re: [Xen-devel] [PATCH MM-PART3 v2 05/12] xen/arm: mm: Introduce _PAGE_PRESENT and _PAGE_POPULATE
On Tue, 14 May 2019, Julien Grall wrote: > At the moment, the flags are not enough to describe what kind of update > will done on the VA range. They need to be used in conjunction with the > enum xenmap_operation. > > It would be more convenient to have all the information for the update > in a single place. > > Two new flags are added to remove the relience on xenmap_operation: > - _PAGE_PRESENT: Indicate whether we are adding/removing the mapping > - _PAGE_POPULATE: Indicate whether we only populate page-tables > > Signed-off-by: Julien Grall <julien.grall@xxxxxxx> > Reviewed-by: Andrii Anisov <andrii_anisov@xxxxxxxx> Looking ahead in this series, I know that this is done so that later on you can remove enum xenmap_operation. But what is the end goal? Why do we want to remove enum xenmap_operation? I guess it is to make the code more reusable? > --- > Changes in v2: > - Add Andrii's reviewed-by > --- > xen/arch/arm/mm.c | 2 +- > xen/include/asm-arm/page.h | 9 +++++++-- > 2 files changed, 8 insertions(+), 3 deletions(-) > > diff --git a/xen/arch/arm/mm.c b/xen/arch/arm/mm.c > index 9de2a1150f..2192dede55 100644 > --- a/xen/arch/arm/mm.c > +++ b/xen/arch/arm/mm.c > @@ -1083,7 +1083,7 @@ int map_pages_to_xen(unsigned long virt, > > int populate_pt_range(unsigned long virt, unsigned long nr_mfns) > { > - return xen_pt_update(RESERVE, virt, INVALID_MFN, nr_mfns, 0); > + return xen_pt_update(RESERVE, virt, INVALID_MFN, nr_mfns, > _PAGE_POPULATE); > } > > int destroy_xen_mappings(unsigned long v, unsigned long e) > diff --git a/xen/include/asm-arm/page.h b/xen/include/asm-arm/page.h > index 2bcdb0f1a5..caf2fac1ff 100644 > --- a/xen/include/asm-arm/page.h > +++ b/xen/include/asm-arm/page.h > @@ -76,6 +76,8 @@ > * > * [0:2] Memory Attribute Index > * [3:4] Permission flags > + * [5] Present bit > + * [6] Populate page table [5] is pretty clear. As a nit, I would probably write: [5] Page Present because there is no need to say "bit", the [5] means it is a bit. Otherwise, something like the following: [5] Present in memory but it's unimportant. It's [6] that we might want to explain a bit better: if we say "Populate page table" then every time we want the Xen pagetable to be populated we would need to pass _PAGE_POPULATE, even when the page is present in memory. Do you see what I mean? I am not entirely sure how to make it clearer. Maybe: [6] Only populate page tables "Only" being the key word. > */ > #define PAGE_AI_MASK(x) ((x) & 0x7U) > > @@ -86,12 +88,15 @@ > #define PAGE_XN_MASK(x) (((x) >> _PAGE_XN_BIT) & 0x1U) > #define PAGE_RO_MASK(x) (((x) >> _PAGE_RO_BIT) & 0x1U) > > +#define _PAGE_PRESENT (1U << 5) > +#define _PAGE_POPULATE (1U << 6) > + > /* > * _PAGE_DEVICE and _PAGE_NORMAL are convenience defines. They are not > * meant to be used outside of this header. > */ > -#define _PAGE_DEVICE _PAGE_XN > -#define _PAGE_NORMAL MT_NORMAL > +#define _PAGE_DEVICE (_PAGE_XN|_PAGE_PRESENT) > +#define _PAGE_NORMAL (MT_NORMAL|_PAGE_PRESENT) > > #define PAGE_HYPERVISOR_RO (_PAGE_NORMAL|_PAGE_RO|_PAGE_XN) > #define PAGE_HYPERVISOR_RX (_PAGE_NORMAL|_PAGE_RO) > -- > 2.11.0 > _______________________________________________ Xen-devel mailing list Xen-devel@xxxxxxxxxxxxxxxxxxxx https://lists.xenproject.org/mailman/listinfo/xen-devel
|
Lists.xenproject.org is hosted with RackSpace, monitoring our |