aboutsummaryrefslogtreecommitdiff
path: root/gcc/graphite-interchange.c
diff options
context:
space:
mode:
authorSebastian Pop <sebastian.pop@amd.com>2009-11-25 05:12:07 +0000
committerSebastian Pop <spop@gcc.gnu.org>2009-11-25 05:12:07 +0000
commit009150e1fc7c05d1f17c2fc4752683436ccd4f5f (patch)
tree52811bc7040b368d537aa797e4e1c3c041ed0206 /gcc/graphite-interchange.c
parent9b196c9394288a3e892456bb3e51d84681de8ff1 (diff)
downloadgcc-009150e1fc7c05d1f17c2fc4752683436ccd4f5f.zip
gcc-009150e1fc7c05d1f17c2fc4752683436ccd4f5f.tar.gz
gcc-009150e1fc7c05d1f17c2fc4752683436ccd4f5f.tar.bz2
graphite-interchange.c: Fix comments.
2009-10-16 Sebastian Pop <sebastian.pop@amd.com> * graphite-interchange.c: Fix comments. * gcc.dg/graphite/block-1.c: Fix comments. From-SVN: r154598
Diffstat (limited to 'gcc/graphite-interchange.c')
-rw-r--r--gcc/graphite-interchange.c3
1 files changed, 2 insertions, 1 deletions
diff --git a/gcc/graphite-interchange.c b/gcc/graphite-interchange.c
index 0e432ce..8926ac3 100644
--- a/gcc/graphite-interchange.c
+++ b/gcc/graphite-interchange.c
@@ -310,7 +310,8 @@ memory_stride_in_loop (Value stride, graphite_dim_t depth, poly_dr_p pdr)
Next, to measure the impact of iterating once in loop "i", we build
a maximization problem: first, we add to DR accesses the dimensions
- k, s2, s3, L1 = 100 * s0 + s1, L2, and D1: polyhedron P1.
+ k, s2, s3, L1 = 100 * s0 + s1, L2, and D1: this is the polyhedron P1.
+ L1 and L2 are the linearized memory access functions.
| i j N a s0 s1 k s2 s3 L1 L2 D1 1
| 0 0 0 1 0 0 0 0 0 0 0 0 -5 = 0 alias = 5