gpu: nvgpu: fix pmu_copy_to_dmem spew

The error check was not taking account of
the DMEM address wrap-around

JIRA DNVGPU-34

Change-Id: Ibfed5532c3ee785b3061e6837f012939118a7ece
Signed-off-by: David Nieto <dmartineznie@nvidia.com>
Reviewed-on: http://git-master/r/1206460
(cherry picked from commit 080953c20f91068ccaaa564d9492a1582ffa28fe)
Reviewed-on: http://git-master/r/1218297
Reviewed-by: Automatic_Commit_Validation_User
GVS: Gerrit_Virtual_Submit
Reviewed-by: Terje Bergstrom <tbergstrom@nvidia.com>
Tested-by: Terje Bergstrom <tbergstrom@nvidia.com>

1 files changed, 1 insertions, 1 deletions

diff --git a/drivers/gpu/nvgpu/gk20a/pmu_gk20a.c b/drivers/gpu/nvgpu/gk20a/pmu_gk20a.c
index 539ad44e..98d6e2df 100644
--- a/drivers/gpu/nvgpu/gk20a/pmu_gk20a.c
+++ b/drivers/gpu/nvgpu/gk20a/pmu_gk20a.c
@@ -2066,7 +2066,7 @@ void pmu_copy_to_dmem(struct pmu_gk20a *pmu,
 
         data = gk20a_readl(g, pwr_falcon_dmemc_r(port)) & addr_mask;
         size = ALIGN(size, 4);
-        if (data != dst + size) {
+        if (data != ((dst + size) & addr_mask)) {
                 gk20a_err(dev_from_gk20a(g),
                         "copy failed. bytes written %d, expected %d",
                         data - dst, size);