diff options
author | xleroy <xleroy@fca1b0fc-160b-0410-b1d3-a4f43f01ea2e> | 2013-07-13 14:02:07 +0000 |
---|---|---|
committer | xleroy <xleroy@fca1b0fc-160b-0410-b1d3-a4f43f01ea2e> | 2013-07-13 14:02:07 +0000 |
commit | f995bde28d1098b51f42a38f3577b903d0420688 (patch) | |
tree | fb0bf1845a3dad1cca50331edebdf05f6864f68d /lib/Floats.v | |
parent | bdac1f6aba5370b21b34c9ee12429c3920b3dffb (diff) | |
download | compcert-kvx-f995bde28d1098b51f42a38f3577b903d0420688.tar.gz compcert-kvx-f995bde28d1098b51f42a38f3577b903d0420688.zip |
More accurate model of condition register flags for ARM and IA32.
git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@2297 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e
Diffstat (limited to 'lib/Floats.v')
-rw-r--r-- | lib/Floats.v | 6 |
1 files changed, 6 insertions, 0 deletions
diff --git a/lib/Floats.v b/lib/Floats.v index 7ae1705c..94c19d25 100644 --- a/lib/Floats.v +++ b/lib/Floats.v @@ -453,6 +453,12 @@ Proof. now apply cmp_le_lt_eq. Qed. +Theorem cmp_lt_gt_false: + forall f1 f2, cmp Clt f1 f2 = true -> cmp Cgt f1 f2 = true -> False. +Proof. + unfold cmp; intros; destruct (order_float f1 f2) as [ [] | ]; discriminate. +Qed. + (** Properties of conversions to/from in-memory representation. The double-precision conversions are bijective (one-to-one). The single-precision conversions lose precision exactly |