- Revised non-overflow constraints on memory injections so that

  injections compose (Values, Memdata, Memory)
- Memory chunks: Mfloat64 now has alignment 8; introduced Mfloat64al32
  that works like old Mfloat64 (i.e. has alignment 4); simplified
  handling of memcpy builtin accordingly.


git-svn-id: https://yquem.inria.fr/compcert/svn/compcert/trunk@1983 fca1b0fc-160b-0410-b1d3-a4f43f01ea2e

5 files changed, 18 insertions, 14 deletions

diff --git a/powerpc/Asm.v b/powerpc/Asm.v
index ea5e4162..5d815fda 100644
--- a/powerpc/Asm.v
+++ b/powerpc/Asm.v
@@ -647,9 +647,9 @@ Definition exec_instr (c: code) (i: instruction) (rs: regset) (m: mem) : outcome
   | Plbzx rd r1 r2 =>
       load2 Mint8unsigned rd r1 r2 rs m
   | Plfd rd cst r1 =>
-      load1 Mfloat64 rd cst r1 rs m
+      load1 Mfloat64al32 rd cst r1 rs m
   | Plfdx rd r1 r2 =>
-      load2 Mfloat64 rd r1 r2 rs m
+      load2 Mfloat64al32 rd r1 r2 rs m
   | Plfs rd cst r1 =>
       load1 Mfloat32 rd cst r1 rs m
   | Plfsx rd r1 r2 =>
@@ -712,9 +712,9 @@ Definition exec_instr (c: code) (i: instruction) (rs: regset) (m: mem) : outcome
   | Pstbx rd r1 r2 =>
       store2 Mint8unsigned rd r1 r2 rs m
   | Pstfd rd cst r1 =>
-      store1 Mfloat64 rd cst r1 rs m
+      store1 Mfloat64al32 rd cst r1 rs m
   | Pstfdx rd r1 r2 =>
-      store2 Mfloat64 rd r1 r2 rs m
+      store2 Mfloat64al32 rd r1 r2 rs m
   | Pstfs rd cst r1 =>
       match store1 Mfloat32 rd cst r1 rs m with
       | OK rs' m' => OK (rs'#FPR13 <- Vundef) m'
@@ -814,7 +814,7 @@ Inductive extcall_arg (rs: regset) (m: mem): loc -> val -> Prop :=
       extcall_arg rs m (S (Outgoing ofs Tint)) v
   | extcall_arg_float_stack: forall ofs bofs v,
       bofs = Stacklayout.fe_ofs_arg + 4 * ofs ->
-      Mem.loadv Mfloat64 m (Val.add (rs (IR GPR1)) (Vint (Int.repr bofs))) = Some v ->
+      Mem.loadv Mfloat64al32 m (Val.add (rs (IR GPR1)) (Vint (Int.repr bofs))) = Some v ->
       extcall_arg rs m (S (Outgoing ofs Tfloat)) v.
 
 Definition extcall_arguments
diff --git a/powerpc/Asmgen.v b/powerpc/Asmgen.v
index f9e4b2cd..b34d9395 100644
--- a/powerpc/Asmgen.v
+++ b/powerpc/Asmgen.v
@@ -485,7 +485,7 @@ Definition transl_instr (f: Mach.function) (i: Mach.instruction) (k: code) :=
       | Mfloat32 =>
           transl_load_store
             (Plfs (freg_of dst)) (Plfsx (freg_of dst)) addr args GPR12 k
-      | Mfloat64 =>
+      | Mfloat64 | Mfloat64al32 =>
           transl_load_store
             (Plfd (freg_of dst)) (Plfdx (freg_of dst)) addr args GPR12 k
       end
@@ -510,7 +510,7 @@ Definition transl_instr (f: Mach.function) (i: Mach.instruction) (k: code) :=
       | Mfloat32 =>
           transl_load_store
             (Pstfs (freg_of src)) (Pstfsx (freg_of src)) addr args temp k
-      | Mfloat64 =>
+      | Mfloat64 | Mfloat64al32 =>
           transl_load_store
             (Pstfd (freg_of src)) (Pstfdx (freg_of src)) addr args temp k
       end
diff --git a/powerpc/Asmgenproof.v b/powerpc/Asmgenproof.v
index e7b73854..e99049c2 100644
--- a/powerpc/Asmgenproof.v
+++ b/powerpc/Asmgenproof.v
@@ -827,9 +827,9 @@ Proof.
     rewrite <- H. apply eval_addressing_preserved. exact symbols_preserved.
   left; eapply exec_straight_steps; eauto with coqlib;
   destruct chunk; simpl; simpl in H6;
-  (* all cases but Mint8signed *)
+  (* all cases but Mint8signed and Mfloat64 *)
   try (eapply transl_load_correct; eauto;
-       intros; simpl; unfold preg_of; rewrite H6; auto).
+       intros; simpl; unfold preg_of; rewrite H6; auto; fail).
   (* Mint8signed *)
   exploit loadv_8_signed_unsigned; eauto. intros [v' [LOAD EQ]].
   assert (X1: forall (cst : constant) (r1 : ireg) (rs1 : regset),
@@ -849,6 +849,10 @@ Proof.
   eapply agree_set_twice_mireg; eauto.
   rewrite EQ. apply Val.sign_ext_lessdef. 
   generalize (ireg_val _ _ _ dst AG1 H6). rewrite Regmap.gss. auto.
+  (* Mfloat64 *)
+  exploit Mem.loadv_float64al32; eauto. intros. clear H0.
+  eapply transl_load_correct; eauto;
+  intros; simpl; unfold preg_of; rewrite H6; auto.
 Qed.
 
 Lemma storev_8_signed_unsigned:
@@ -883,6 +887,7 @@ Proof.
   rewrite <- (eval_addressing_preserved _ _ symbols_preserved) in H.
   left; eapply exec_straight_steps_bis; eauto with coqlib.
   destruct chunk; simpl; simpl in H6;
+  try (generalize (Mem.storev_float64al32 _ _ _ _ H0); intros);
   try (rewrite storev_8_signed_unsigned in H0);
   try (rewrite storev_16_signed_unsigned in H0);
   simpl; eapply transl_store_correct; eauto;
diff --git a/powerpc/PrintAsm.ml b/powerpc/PrintAsm.ml
index 3331a1c2..ac9a5da4 100644
--- a/powerpc/PrintAsm.ml
+++ b/powerpc/PrintAsm.ml
@@ -350,7 +350,7 @@ let print_builtin_vload_common oc chunk base offset res =
       fprintf oc "	lwz	%a, %a(%a)\n" ireg res constant offset ireg base
   | Mfloat32, FR res ->
       fprintf oc "	lfs	%a, %a(%a)\n" freg res constant offset ireg base
-  | Mfloat64, FR res ->
+  | (Mfloat64 | Mfloat64al32), FR res ->
       fprintf oc "	lfd	%a, %a(%a)\n" freg res constant offset ireg base
   | _ ->
       assert false
@@ -383,7 +383,7 @@ let print_builtin_vstore_common oc chunk base offset src =
   | Mfloat32, FR src ->
       fprintf oc "	frsp	%a, %a\n" freg FPR13 freg src;
       fprintf oc "	stfs	%a, %a(%a)\n" freg FPR13 constant offset ireg base
-  | Mfloat64, FR src ->
+  | (Mfloat64 | Mfloat64al32), FR src ->
       fprintf oc "	stfd	%a, %a(%a)\n" freg src constant offset ireg base
   | _ ->
       assert false
diff --git a/powerpc/eabi/Stacklayout.v b/powerpc/eabi/Stacklayout.v
index 22a28269..cd4d9ae4 100644
--- a/powerpc/eabi/Stacklayout.v
+++ b/powerpc/eabi/Stacklayout.v
@@ -113,7 +113,7 @@ Remark frame_env_aligned:
   /\ (8 | fe.(fe_ofs_float_local))
   /\ (8 | fe.(fe_ofs_float_callee_save))
   /\ (4 | fe.(fe_ofs_retaddr))
-  /\ (4 | fe.(fe_stack_data))
+  /\ (8 | fe.(fe_stack_data))
   /\ (16 | fe.(fe_size)).
 Proof.
   intros.
@@ -133,8 +133,7 @@ Proof.
   set (x5 := x4 + 8 * bound_float_local b).
   assert (8 | x5). unfold x5. apply Zdivide_plus_r; auto. exists (bound_float_local b); ring.
   set (x6 := x5 + 8 * bound_float_callee_save b).
-  assert (4 | x6).
-    apply Zdivides_trans with 8. exists 2; auto.
+  assert (8 | x6).
     unfold x6. apply Zdivide_plus_r; auto. exists (bound_float_callee_save b); ring.
   set (x7 := align (x6 + bound_stack_data b) 16).
   assert (16 | x7). unfold x7; apply align_divides. omega.