gpu: nvgpu: fix issue with DIV_ROUND_CLOSEST

Fix the issue in DIV_ROUND_CLOSEST implementation. Current
implementation will have issues for an input of two negative
numbers. For example, Input values of a = -13 and divisor = -5
should ideally give result as 3. But because of the third condition
check ie.(val) > 0, the result would be given incorrectly as 2
with the current implementation. Modify the condition to handle
two negative number division.

Jira NVGPU-4478

Change-Id: Ifb3f58b4ff742abbb59fda328c1b5facdcb1b38b
Signed-off-by: ajesh <akv@nvidia.com>
Reviewed-on: https://git-master.nvidia.com/r/c/linux-nvgpu/+/2284320
Reviewed-by: Automatic_Commit_Validation_User
Reviewed-by: Alex Waterman <alexw@nvidia.com>
Reviewed-by: mobile promotions <svcmobile_promotions@nvidia.com>
Tested-by: mobile promotions <svcmobile_promotions@nvidia.com>
GVS: Gerrit_Virtual_Submit

drivers/gpu/nvgpu/include/nvgpu/posix/utils.h

@@ -146,7 +146,8 @@
 	typeof(a) val = (a);                            \
 	typeof(divisor) div = (divisor);                \
 	(((typeof(a))-1) > 0 ||                         \
-	((typeof(divisor))-1) > 0 || (val) > 0) ?       \
+	((typeof(divisor))-1) > 0 ||                    \
+	(((val) > 0) == ((div) > 0))) ?                 \
 		(((val) + ((div) / 2)) / (div)) :       \
 		(((val) - ((div) / 2)) / (div));        \
 }                                                       \