gpu: nvgpu: Apply GM20B dfs_det_cal mask

Applied dfs_det_cal field mask when updating DFS external calibration
setting to prevent overflow for negative setting (not happening so far
with ADC offset always below Vmin, but may be needed if/when ADC fusing
changes).

Change-Id: I94b415edec84111b957b73ccfc80c1a8640b8e3a
Signed-off-by: Alex Frid <afrid@nvidia.com>
Reviewed-on: http://git-master/r/1504242
(cherry picked from commit 9b8072ffcc02f6684a56609c25d1d77ed0852f7a)
Reviewed-on: https://git-master.nvidia.com/r/1525831
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>

1 files changed, 1 insertions, 1 deletions

diff --git a/drivers/gpu/nvgpu/gm20b/clk_gm20b.c b/drivers/gpu/nvgpu/gm20b/clk_gm20b.c
index addc27bb..8f770e2e 100644
--- a/drivers/gpu/nvgpu/gm20b/clk_gm20b.c
+++ b/drivers/gpu/nvgpu/gm20b/clk_gm20b.c
@@ -416,7 +416,7 @@ static void clk_set_dfs_ext_cal(struct gk20a *g, u32 dfs_det_cal)
 
         data = gk20a_readl(g, trim_gpc_bcast_gpcpll_dvfs2_r());
         data &= ~(BIT(DFS_DET_RANGE + 1) - 1);
-        data |= dfs_det_cal;
+        data |= dfs_det_cal & (BIT(DFS_DET_RANGE + 1) - 1);
         gk20a_writel(g, trim_gpc_bcast_gpcpll_dvfs2_r(), data);
 
         data = gk20a_readl(g, trim_sys_gpcpll_dvfs1_r());