Some architecture ports like ARC don't provide the PHYS_OFFSET symbol. Define it to 0 in that case, which is the most conservative default in the usage context of the etnaviv driver.
Signed-off-by: Lucas Stach l.stach@pengutronix.de --- drivers/gpu/drm/etnaviv/etnaviv_gpu.c | 4 ++++ 1 file changed, 4 insertions(+)
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c index 935d99be748e..febbd1e5bbc7 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c @@ -31,6 +31,10 @@ #include "state_hi.xml.h" #include "cmdstream.xml.h"
+#ifndef PHYS_OFFSET +#define PHYS_OFFSET 0 +#endif + static const struct platform_device_id gpu_ids[] = { { .name = "etnaviv-gpu,2d" }, { },
It seems wrong to define PHYS_OFFSET here in a gpu driver. Is there an actual ARC system using vivante? The bit of code that uses PHYS_OFFSET looks pretty ARM specific.
If not, then at least maybe
+#if !define (PHYS_OFFSET) && defined (COMPILE_TEST) +#define PHY_OFFSET 0 +#endif
would be more appropriate?
Maybe the use of PHYS_OFFSET should be moved into an inline in a header where it can be defined differently for other arches
On Sun, Jan 7, 2018 at 7:56 AM, Lucas Stach l.stach@pengutronix.de wrote:
Some architecture ports like ARC don't provide the PHYS_OFFSET symbol. Define it to 0 in that case, which is the most conservative default in the usage context of the etnaviv driver.
Signed-off-by: Lucas Stach l.stach@pengutronix.de
drivers/gpu/drm/etnaviv/etnaviv_gpu.c | 4 ++++ 1 file changed, 4 insertions(+)
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c index 935d99be748e..febbd1e5bbc7 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c @@ -31,6 +31,10 @@ #include "state_hi.xml.h" #include "cmdstream.xml.h"
+#ifndef PHYS_OFFSET +#define PHYS_OFFSET 0 +#endif
static const struct platform_device_id gpu_ids[] = { { .name = "etnaviv-gpu,2d" }, { }, -- 2.11.0
etnaviv mailing list etnaviv@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/etnaviv
Am Montag, den 08.01.2018, 11:16 -0700 schrieb Joshua Clayton:
It seems wrong to define PHYS_OFFSET here in a gpu driver. Is there an actual ARC system using vivante?
Yes.
The bit of code that uses PHYS_OFFSET looks pretty ARM specific.
If not, then at least maybe
+#if !define (PHYS_OFFSET) && defined (COMPILE_TEST) +#define PHY_OFFSET 0 +#endif
would be more appropriate?
Maybe the use of PHYS_OFFSET should be moved into an inline in a header where it can be defined differently for other arches
I don't want to spread the symbol around. It is defined to some (etnaviv specific) safe value in exactly the compilation unit where it is used.
Regards, Lucas
On Sun, Jan 7, 2018 at 7:56 AM, Lucas Stach l.stach@pengutronix.de wrote:
Some architecture ports like ARC don't provide the PHYS_OFFSET symbol. Define it to 0 in that case, which is the most conservative default in the usage context of the etnaviv driver.
Signed-off-by: Lucas Stach l.stach@pengutronix.de
drivers/gpu/drm/etnaviv/etnaviv_gpu.c | 4 ++++ 1 file changed, 4 insertions(+)
diff --git a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c index 935d99be748e..febbd1e5bbc7 100644 --- a/drivers/gpu/drm/etnaviv/etnaviv_gpu.c +++ b/drivers/gpu/drm/etnaviv/etnaviv_gpu.c @@ -31,6 +31,10 @@ #include "state_hi.xml.h" #include "cmdstream.xml.h"
+#ifndef PHYS_OFFSET +#define PHYS_OFFSET 0 +#endif
static const struct platform_device_id gpu_ids[] = { { .name = "etnaviv-gpu,2d" }, { }, -- 2.11.0
etnaviv mailing list etnaviv@lists.freedesktop.org https://lists.freedesktop.org/mailman/listinfo/etnaviv
dri-devel@lists.freedesktop.org
-
Joshua Clayton -
Lucas Stach