X-Git-Url: https://git.distorted.org.uk/~mdw/catacomb/blobdiff_plain/92edc356a312bc64abca0c30bc03d4b6676f3d39..6966e7a60a87415d3d02230608a98016c03a7a51:/math/mpx-mul4-amd64-sse2.S?ds=sidebyside

diff --git a/math/mpx-mul4-amd64-sse2.S b/math/mpx-mul4-amd64-sse2.S
index a37aba69..17c4f1ad 100644
--- a/math/mpx-mul4-amd64-sse2.S
+++ b/math/mpx-mul4-amd64-sse2.S
@@ -1260,7 +1260,9 @@ FUNC(mpxmont_redc4_amd64_sse2)
 	cmp	rdi, DVL4
 	jb	0b
 
-	// Deal with the tail end.
+	// Deal with the tail end.  Note that the actual destination length
+	// won't be an exacty number of blocks of four, so it's safe to just
+	// drop through here.
 7:	add	[rdi], C
 	mov	C, 0
 	add	rdi, 4
@@ -1268,9 +1270,7 @@ FUNC(mpxmont_redc4_amd64_sse2)
 	cmp	rdi, DVL
 	jb	7b
 
-	// All done for this iteration.  Start the next.  (This must have at
-	// least one follow-on iteration, or we'd not have started this outer
-	// loop.)
+	// All done for this iteration.  Start the next.
 8:	mov	rdi, DV			// -> Z = dv[i]
 	mov	rbx, NV			// -> X = nv[0]
 	cmp	rdi, DVLO		// all done yet?