io-mapping: ensure io_mapping_map_atomic _is_ atomic
For the !HAVE_ATOMIC_IOMAP case the stub functions did not call
pagefault_disable/_enable. The i915 driver relies on the map
actually being atomic, otherwise it can deadlock with it's own
pagefault handler in the gtt pwrite fastpath.
This is exercised by gem_mmap_gtt from the intel-gpu-toosl gem
testsuite.
v2: Chris Wilson noted the lack of an include.
Bugzilla: https://bugs.freedesktop.org/show_bug.cgi?id=38115
Cc: stable@kernel.org
Signed-off-by: Daniel Vetter <daniel.vetter@ffwll.ch>
Reviewed-by: Chris Wilson <chris@chris-wilson.co.uk>
Signed-off-by: Keith Packard <keithp@keithp.com>
diff --git a/include/linux/io-mapping.h b/include/linux/io-mapping.h
index 8cdcc2a1..1feeb52 100644
--- a/include/linux/io-mapping.h
+++ b/include/linux/io-mapping.h
@@ -117,6 +117,8 @@
#else
+#include <linux/uaccess.h>
+
/* this struct isn't actually defined anywhere */
struct io_mapping;
@@ -138,12 +140,14 @@
io_mapping_map_atomic_wc(struct io_mapping *mapping,
unsigned long offset)
{
+ pagefault_disable();
return ((char __force __iomem *) mapping) + offset;
}
static inline void
io_mapping_unmap_atomic(void __iomem *vaddr)
{
+ pagefault_enable();
}
/* Non-atomic map/unmap */