I thought this was going to be a very simple small bugfix for Michał's
Processor Trace series. Serves me right for expecting it not to be full of
bear traps...
The sole implementation of acquire_resource never asks for size, so its little
surprise that Xen is broken for compat callers, and returns incorrect
information for regular callers.
v2 was delayed substantially due to the discovery of XSA-334, but is complete
now, permitting the mapping of arbitrary sized resouces, along with fixes to
the compat XLAT logic.
v3 was delayed substanitally due to other security work. The major change
from v2 is how the size request works from userspace, which now depends on:
https://lore.kernel.org/xen-devel/20210112115358.23346-1-roger.pau@citrix.com/T/#u
to fix the IOCTL in Linux without breaking ioctl-restrict usecases.
A branch can be obtained from:
https://xenbits.xen.org/gitweb/?p=people/andrewcoop/xen.git;a=shortlog;h=refs/heads/xen-acquire-resource
Andrew Cooper (7):
xen/gnttab: Rework resource acquisition
xen/memory: Fix acquire_resource size semantics
tools/foreignmem: Support querying the size of a resource
xen/memory: Clarify the XENMEM_acquire_resource ABI description
xen/memory: Improve compat XENMEM_acquire_resource handling
xen/memory: Indent part of acquire_resource()
xen/memory: Fix mapping grant tables with XENMEM_acquire_resource
tools/include/xenforeignmemory.h | 15 ++
tools/libs/foreignmemory/Makefile | 2 +-
tools/libs/foreignmemory/core.c | 18 ++
tools/libs/foreignmemory/freebsd.c | 18 +-
tools/libs/foreignmemory/libxenforeignmemory.map | 4 +
tools/libs/foreignmemory/linux.c | 18 +-
xen/arch/x86/mm.c | 24 ++-
xen/common/compat/memory.c | 147 ++++++++++++----
xen/common/grant_table.c | 106 +++++++++---
xen/common/memory.c | 210 ++++++++++++++---------
xen/include/asm-x86/mm.h | 3 +
xen/include/public/memory.h | 23 ++-
xen/include/xen/grant_table.h | 21 ++-
xen/include/xen/mm.h | 6 +
14 files changed, 445 insertions(+), 170 deletions(-)
--
2.11.0