Skip to contentSkip to conversation

ppc: Add support for 'mffsce' instruction

PATCH In review Paul A. Clarke 3 participants 3 messages Sep 12, 2019 – Sep 16, 2019

lore
Share & tools
lore
#0 lore raw

PATCH 1/1 ppc: Add support for 'mffsce' instruction

From: "Paul A. Clarke" <[email protected]>

ISA 3.0B added a set of Floating-Point Status and Control Register (FPSCR)
instructions: mffsce, mffscdrn, mffscdrni, mffscrn, mffscrni, mffsl. This patch adds support for 'mffsce' instruction.

'mffsce' is identical to 'mffs', except that it also clears the exception enable bits in the FPSCR.

On CPUs without support for 'mffsce' (below ISA 3.0), the instruction will execute identically to 'mffs'.

---
 target/ppc/translate/fp-impl.inc.c | 30 ++++++++++++++++++++++++++++++
 target/ppc/translate/fp-ops.inc.c  |  2 ++
 2 files changed, 32 insertions(+)
target/ppc/translate/fp-impl.inc.c +30 -0
index 59a4faf..34edc45 100644
@@ -639,6 +639,36 @@ static void gen_mffsl(DisasContext *ctx)
     tcg_temp_free_i64(t0);
 }
 
+/* mffsce */
+static void gen_mffsce(DisasContext *ctx)
+{
+    TCGv_i64 t0;
+    TCGv_i32 mask;
+
+    if (unlikely(!(ctx->insns_flags2 & PPC2_ISA300))) {
+        return gen_mffs(ctx);
+    }
+
+    if (unlikely(!ctx->fpu_enabled)) {
+        gen_exception(ctx, POWERPC_EXCP_FPU);
+        return;
+    }
+
+    t0 = tcg_temp_new_i64();
+
+    gen_reset_fpstatus();
+    tcg_gen_extu_tl_i64(t0, cpu_fpscr);
+    set_fpr(rD(ctx->opcode), t0);
+
+    /* Clear exception enable bits in the FPSCR.  */
+    tcg_gen_andi_i64(t0, t0, ~FP_ENABLES);
+    mask = tcg_const_i32(0x0003);
+    gen_helper_store_fpscr(cpu_env, t0, mask);
+
+    tcg_temp_free_i32(mask);
+    tcg_temp_free_i64(t0);
+}
+
 static void gen_helper_mffscrn(DisasContext *ctx, TCGv_i64 t1)
 {
     TCGv_i64 t0 = tcg_temp_new_i64();
target/ppc/translate/fp-ops.inc.c +2 -0
index f2bcf0e..88fab65 100644
@@ -105,6 +105,8 @@ GEN_HANDLER_E(fmrgew, 0x3F, 0x06, 0x1E, 0x00000001, PPC_NONE, PPC2_VSX207),
 GEN_HANDLER_E(fmrgow, 0x3F, 0x06, 0x1A, 0x00000001, PPC_NONE, PPC2_VSX207),
 GEN_HANDLER(mcrfs, 0x3F, 0x00, 0x02, 0x0063F801, PPC_FLOAT),
 GEN_HANDLER_E_2(mffs, 0x3F, 0x07, 0x12, 0x00, 0x00000000, PPC_FLOAT, PPC_NONE),
+GEN_HANDLER_E_2(mffsce, 0x3F, 0x07, 0x12, 0x01, 0x00000000, PPC_FLOAT,
+    PPC2_ISA300),
 GEN_HANDLER_E_2(mffsl, 0x3F, 0x07, 0x12, 0x18, 0x00000000, PPC_FLOAT,
     PPC2_ISA300),
 GEN_HANDLER_E_2(mffscrn, 0x3F, 0x07, 0x12, 0x16, 0x00000000, PPC_FLOAT,
Signature · 1 line
1.8.3.1
Reviewed-by on v1 1/1
#1 lore raw
Quoted 34 lines · Paul A. Clarke

From: "Paul A. Clarke" <[email protected]>

ISA 3.0B added a set of Floating-Point Status and Control Register (FPSCR)
instructions: mffsce, mffscdrn, mffscdrni, mffscrn, mffscrni, mffsl. This patch adds support for 'mffsce' instruction.

'mffsce' is identical to 'mffs', except that it also clears the exception enable bits in the FPSCR.

On CPUs without support for 'mffsce' (below ISA 3.0), the instruction will execute identically to 'mffs'.

---
 target/ppc/translate/fp-impl.inc.c | 30 ++++++++++++++++++++++++++++++
 target/ppc/translate/fp-ops.inc.c  |  2 ++
 2 files changed, 32 insertions(+)
target/ppc/translate/fp-impl.inc.c +8 -0
index 59a4faf..34edc45 100644
@@ -639,6 +639,36 @@ static void gen_mffsl(DisasContext *ctx)
     tcg_temp_free_i64(t0);
 }
 
+/* mffsce */
+static void gen_mffsce(DisasContext *ctx)
+{
+    TCGv_i64 t0;
+    TCGv_i32 mask;
+
+    if (unlikely(!(ctx->insns_flags2 & PPC2_ISA300))) {
+        return gen_mffs(ctx);

target/ppc/translate/fp-impl.inc.c static void gen_mffsl(DisasContext *ctx) · line 659

Quoted 10 lines
target/ppc/translate/fp-impl.inc.c +10 -0
+    }
+
+    if (unlikely(!ctx->fpu_enabled)) {
+        gen_exception(ctx, POWERPC_EXCP_FPU);
+        return;
+    }
+
+    t0 = tcg_temp_new_i64();
+
+    gen_reset_fpstatus();

Note for future cleanup: we should not need to sprinkle these all over. This should be the steady-state condition after softfp exceptions have been processed into powerpc exceptions, after every single fp instruction.

That said, you're mirroring gen_mffs here, and the cleanup should happen globally.

r~

#2 lore raw
Quoted 50 lines · Richard Henderson
Quoted 34 lines · Paul A. Clarke

From: "Paul A. Clarke" <[email protected]>

ISA 3.0B added a set of Floating-Point Status and Control Register (FPSCR)
instructions: mffsce, mffscdrn, mffscdrni, mffscrn, mffscrni, mffsl. This patch adds support for 'mffsce' instruction.

'mffsce' is identical to 'mffs', except that it also clears the exception enable bits in the FPSCR.

On CPUs without support for 'mffsce' (below ISA 3.0), the instruction will execute identically to 'mffs'.

---
 target/ppc/translate/fp-impl.inc.c | 30 ++++++++++++++++++++++++++++++
 target/ppc/translate/fp-ops.inc.c  |  2 ++
 2 files changed, 32 insertions(+)
target/ppc/translate/fp-impl.inc.c +8 -0
index 59a4faf..34edc45 100644
@@ -639,6 +639,36 @@ static void gen_mffsl(DisasContext *ctx)
     tcg_temp_free_i64(t0);
 }
 
+/* mffsce */
+static void gen_mffsce(DisasContext *ctx)
+{
+    TCGv_i64 t0;
+    TCGv_i32 mask;
+
+    if (unlikely(!(ctx->insns_flags2 & PPC2_ISA300))) {
+        return gen_mffs(ctx);

target/ppc/translate/fp-impl.inc.c static void gen_mffsl(DisasContext *ctx) · line 659

Quoted 10 lines
target/ppc/translate/fp-impl.inc.c +10 -0
+    }
+
+    if (unlikely(!ctx->fpu_enabled)) {
+        gen_exception(ctx, POWERPC_EXCP_FPU);
+        return;
+    }
+
+    t0 = tcg_temp_new_i64();
+
+    gen_reset_fpstatus();

Note for future cleanup: we should not need to sprinkle these all over. This should be the steady-state condition after softfp exceptions have been processed into powerpc exceptions, after every single fp instruction.

Quoted 2 lines

That said, you're mirroring gen_mffs here, and the cleanup should happen globally.

This doesn't apply, presuambly because it's on top of the other patch. So I'll wait for the respin that includes the small update that rtc requested on that one.

Quoted 6 lines

r~

Signature · 4 lines
David Gibson			| I'll have my music baroque, and my code
david AT gibson.dropbear.id.au	| minimalist, thank you.  NOT _the_ _other_
				| _way_ _around_!
http://www.ozlabs.org/~dgibson