* [PATCH v3 0/4] target/sh4: Fix ADDV/SUBV opcodes
@ 2024-04-30 14:56 Philippe Mathieu-Daudé
2024-04-30 14:56 ` [PATCH v3 1/4] target/sh4: Fix ADDV opcode Philippe Mathieu-Daudé
` (4 more replies)
0 siblings, 5 replies; 6+ messages in thread
From: Philippe Mathieu-Daudé @ 2024-04-30 14:56 UTC (permalink / raw)
To: qemu-devel
Cc: Yoshinori Sato, Paul Cercueil, John Paul Adrian Glaubitz,
Philippe =?unknown-8bit?q?Mathieu-Daud=C3=A9?=
Since v2:
- Add tests (Paul)
- Rename TCGv variables as in manual
Philippe Mathieu-Daudé (4):
target/sh4: Fix ADDV opcode
target/sh4: Fix SUBV opcode
target/sh4: Rename TCGv variables as manual for ADDV opcode
target/sh4: Rename TCGv variables as manual for SUBV opcode
target/sh4/translate.c | 32 ++++++++++++++++++++------------
tests/tcg/sh4/test-addv.c | 23 +++++++++++++++++++++++
tests/tcg/sh4/test-subv.c | 26 ++++++++++++++++++++++++++
tests/tcg/sh4/Makefile.target | 6 ++++++
4 files changed, 75 insertions(+), 12 deletions(-)
create mode 100644 tests/tcg/sh4/test-addv.c
create mode 100644 tests/tcg/sh4/test-subv.c
--
2.41.0
^ permalink raw reply [flat|nested] 6+ messages in thread
* [PATCH v3 1/4] target/sh4: Fix ADDV opcode
2024-04-30 14:56 [PATCH v3 0/4] target/sh4: Fix ADDV/SUBV opcodes Philippe Mathieu-Daudé
@ 2024-04-30 14:56 ` Philippe Mathieu-Daudé
2024-04-30 14:56 ` [PATCH v3 2/4] target/sh4: Fix SUBV opcode Philippe Mathieu-Daudé
` (3 subsequent siblings)
4 siblings, 0 replies; 6+ messages in thread
From: Philippe Mathieu-Daudé @ 2024-04-30 14:56 UTC (permalink / raw)
To: qemu-devel
Cc: Yoshinori Sato, Paul Cercueil, John Paul Adrian Glaubitz,
Philippe Mathieu-Daudé, qemu-stable
The documentation says:
ADDV Rm, Rn Rn + Rm -> Rn, overflow -> T
But QEMU implementation was:
ADDV Rm, Rn Rn + Rm -> Rm, overflow -> T
Fix by filling the correct Rm register.
Add tests provided by Paul Cercueil.
Cc: qemu-stable@nongnu.org
Fixes: ad8d25a11f ("target-sh4: implement addv and subv using TCG")
Reported-by: Paul Cercueil <paul@crapouillou.net>
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/2317
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
---
target/sh4/translate.c | 2 +-
tests/tcg/sh4/test-addv.c | 23 +++++++++++++++++++++++
tests/tcg/sh4/Makefile.target | 3 +++
3 files changed, 27 insertions(+), 1 deletion(-)
create mode 100644 tests/tcg/sh4/test-addv.c
diff --git a/target/sh4/translate.c b/target/sh4/translate.c
index ebb6c901bf..4a1dd0d1f4 100644
--- a/target/sh4/translate.c
+++ b/target/sh4/translate.c
@@ -714,7 +714,7 @@ static void _decode_opc(DisasContext * ctx)
tcg_gen_xor_i32(t2, REG(B7_4), REG(B11_8));
tcg_gen_andc_i32(cpu_sr_t, t1, t2);
tcg_gen_shri_i32(cpu_sr_t, cpu_sr_t, 31);
- tcg_gen_mov_i32(REG(B7_4), t0);
+ tcg_gen_mov_i32(REG(B11_8), t0);
}
return;
case 0x2009: /* and Rm,Rn */
diff --git a/tests/tcg/sh4/test-addv.c b/tests/tcg/sh4/test-addv.c
new file mode 100644
index 0000000000..54ac77b867
--- /dev/null
+++ b/tests/tcg/sh4/test-addv.c
@@ -0,0 +1,23 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+
+#include <assert.h>
+#include <limits.h>
+
+static void addv(int a, int b, int res, int carry)
+{
+ unsigned int c;
+
+ asm volatile("addv %2,%0\n"
+ "movt %1\n"
+ : "+r"(a), "=r"(c) : "r"(b) :);
+
+ assert(c == carry && a == res);
+}
+
+int main(void)
+{
+ addv(INT_MAX, 1, INT_MIN, 1);
+ addv(INT_MAX - 1, 1, INT_MAX, 0);
+
+ return 0;
+}
diff --git a/tests/tcg/sh4/Makefile.target b/tests/tcg/sh4/Makefile.target
index 4d09291c0c..521b8b0a76 100644
--- a/tests/tcg/sh4/Makefile.target
+++ b/tests/tcg/sh4/Makefile.target
@@ -17,3 +17,6 @@ TESTS += test-macl
test-macw: CFLAGS += -O -g
TESTS += test-macw
+
+test-addv: CFLAGS += -O -g
+TESTS += test-addv
--
2.41.0
^ permalink raw reply related [flat|nested] 6+ messages in thread
* [PATCH v3 2/4] target/sh4: Fix SUBV opcode
2024-04-30 14:56 [PATCH v3 0/4] target/sh4: Fix ADDV/SUBV opcodes Philippe Mathieu-Daudé
2024-04-30 14:56 ` [PATCH v3 1/4] target/sh4: Fix ADDV opcode Philippe Mathieu-Daudé
@ 2024-04-30 14:56 ` Philippe Mathieu-Daudé
2024-04-30 14:56 ` [PATCH v3 3/4] target/sh4: Rename TCGv variables as manual for ADDV opcode Philippe Mathieu-Daudé
` (2 subsequent siblings)
4 siblings, 0 replies; 6+ messages in thread
From: Philippe Mathieu-Daudé @ 2024-04-30 14:56 UTC (permalink / raw)
To: qemu-devel
Cc: Yoshinori Sato, Paul Cercueil, John Paul Adrian Glaubitz,
Philippe Mathieu-Daudé, qemu-stable
The documentation says:
SUBV Rm, Rn Rn - Rm -> Rn, underflow -> T
The overflow / underflow can be calculated as:
T = ((Rn ^ Rm) & (Result ^ Rn)) >> 31
However we were using the incorrect:
T = ((Rn ^ Rm) & (Result ^ Rm)) >> 31
Fix by using the Rn register instead of Rm.
Add tests provided by Paul Cercueil.
Cc: qemu-stable@nongnu.org
Fixes: ad8d25a11f ("target-sh4: implement addv and subv using TCG")
Reported-by: Paul Cercueil <paul@crapouillou.net>
Suggested-by: Paul Cercueil <paul@crapouillou.net>
Resolves: https://gitlab.com/qemu-project/qemu/-/issues/2318
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
---
target/sh4/translate.c | 2 +-
tests/tcg/sh4/test-subv.c | 26 ++++++++++++++++++++++++++
tests/tcg/sh4/Makefile.target | 3 +++
3 files changed, 30 insertions(+), 1 deletion(-)
create mode 100644 tests/tcg/sh4/test-subv.c
diff --git a/target/sh4/translate.c b/target/sh4/translate.c
index 4a1dd0d1f4..3e013b7c7c 100644
--- a/target/sh4/translate.c
+++ b/target/sh4/translate.c
@@ -933,7 +933,7 @@ static void _decode_opc(DisasContext * ctx)
t0 = tcg_temp_new();
tcg_gen_sub_i32(t0, REG(B11_8), REG(B7_4));
t1 = tcg_temp_new();
- tcg_gen_xor_i32(t1, t0, REG(B7_4));
+ tcg_gen_xor_i32(t1, t0, REG(B11_8));
t2 = tcg_temp_new();
tcg_gen_xor_i32(t2, REG(B11_8), REG(B7_4));
tcg_gen_and_i32(t1, t1, t2);
diff --git a/tests/tcg/sh4/test-subv.c b/tests/tcg/sh4/test-subv.c
new file mode 100644
index 0000000000..d28a9f8f89
--- /dev/null
+++ b/tests/tcg/sh4/test-subv.c
@@ -0,0 +1,26 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+
+#include <assert.h>
+#include <limits.h>
+
+static void subv(int a, int b, int res, int carry)
+{
+ unsigned int c;
+
+ asm volatile("subv %2,%0\n"
+ "movt %1\n"
+ : "+r"(a), "=r"(c) : "r"(b) :);
+
+ assert(c == carry && a == res);
+}
+
+int main(void)
+{
+ subv(INT_MIN, 1, INT_MAX, 1);
+ subv(INT_MAX, -1, INT_MIN, 1);
+ subv(INT_MAX, 1, INT_MAX - 1, 0);
+ subv(0, 1, -1, 0);
+ subv(-1, -1, 0, 0);
+
+ return 0;
+}
diff --git a/tests/tcg/sh4/Makefile.target b/tests/tcg/sh4/Makefile.target
index 521b8b0a76..7852fa62d8 100644
--- a/tests/tcg/sh4/Makefile.target
+++ b/tests/tcg/sh4/Makefile.target
@@ -20,3 +20,6 @@ TESTS += test-macw
test-addv: CFLAGS += -O -g
TESTS += test-addv
+
+test-subv: CFLAGS += -O -g
+TESTS += test-subv
--
2.41.0
^ permalink raw reply related [flat|nested] 6+ messages in thread
* [PATCH v3 3/4] target/sh4: Rename TCGv variables as manual for ADDV opcode
2024-04-30 14:56 [PATCH v3 0/4] target/sh4: Fix ADDV/SUBV opcodes Philippe Mathieu-Daudé
2024-04-30 14:56 ` [PATCH v3 1/4] target/sh4: Fix ADDV opcode Philippe Mathieu-Daudé
2024-04-30 14:56 ` [PATCH v3 2/4] target/sh4: Fix SUBV opcode Philippe Mathieu-Daudé
@ 2024-04-30 14:56 ` Philippe Mathieu-Daudé
2024-04-30 14:56 ` [PATCH v3 4/4] target/sh4: Rename TCGv variables as manual for SUBV opcode Philippe Mathieu-Daudé
2024-04-30 15:42 ` [PATCH v3 0/4] target/sh4: Fix ADDV/SUBV opcodes Richard Henderson
4 siblings, 0 replies; 6+ messages in thread
From: Philippe Mathieu-Daudé @ 2024-04-30 14:56 UTC (permalink / raw)
To: qemu-devel
Cc: Yoshinori Sato, Paul Cercueil, John Paul Adrian Glaubitz,
Philippe Mathieu-Daudé
To easily compare with the SH4 manual, rename:
REG(B11_8) -> Rn
REG(B7_4) -> Rm
t0 -> result
Mention how overflow is calculated.
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
---
target/sh4/translate.c | 16 ++++++++++------
1 file changed, 10 insertions(+), 6 deletions(-)
diff --git a/target/sh4/translate.c b/target/sh4/translate.c
index 3e013b7c7c..47c0f3404e 100644
--- a/target/sh4/translate.c
+++ b/target/sh4/translate.c
@@ -705,16 +705,20 @@ static void _decode_opc(DisasContext * ctx)
return;
case 0x300f: /* addv Rm,Rn */
{
- TCGv t0, t1, t2;
- t0 = tcg_temp_new();
- tcg_gen_add_i32(t0, REG(B7_4), REG(B11_8));
+ TCGv Rn = REG(B11_8);
+ TCGv Rm = REG(B7_4);
+ TCGv result, t1, t2;
+
+ result = tcg_temp_new();
t1 = tcg_temp_new();
- tcg_gen_xor_i32(t1, t0, REG(B11_8));
t2 = tcg_temp_new();
- tcg_gen_xor_i32(t2, REG(B7_4), REG(B11_8));
+ tcg_gen_add_i32(result, Rm, Rn);
+ /* T = ((Rn ^ Rm) & (Result ^ Rn)) >> 31 */
+ tcg_gen_xor_i32(t1, result, Rn);
+ tcg_gen_xor_i32(t2, Rm, Rn);
tcg_gen_andc_i32(cpu_sr_t, t1, t2);
tcg_gen_shri_i32(cpu_sr_t, cpu_sr_t, 31);
- tcg_gen_mov_i32(REG(B11_8), t0);
+ tcg_gen_mov_i32(Rn, result);
}
return;
case 0x2009: /* and Rm,Rn */
--
2.41.0
^ permalink raw reply related [flat|nested] 6+ messages in thread
* [PATCH v3 4/4] target/sh4: Rename TCGv variables as manual for SUBV opcode
2024-04-30 14:56 [PATCH v3 0/4] target/sh4: Fix ADDV/SUBV opcodes Philippe Mathieu-Daudé
` (2 preceding siblings ...)
2024-04-30 14:56 ` [PATCH v3 3/4] target/sh4: Rename TCGv variables as manual for ADDV opcode Philippe Mathieu-Daudé
@ 2024-04-30 14:56 ` Philippe Mathieu-Daudé
2024-04-30 15:42 ` [PATCH v3 0/4] target/sh4: Fix ADDV/SUBV opcodes Richard Henderson
4 siblings, 0 replies; 6+ messages in thread
From: Philippe Mathieu-Daudé @ 2024-04-30 14:56 UTC (permalink / raw)
To: qemu-devel
Cc: Yoshinori Sato, Paul Cercueil, John Paul Adrian Glaubitz,
Philippe Mathieu-Daudé
To easily compare with the SH4 manual, rename:
REG(B11_8) -> Rn
REG(B7_4) -> Rm
t0 -> result
Mention how underflow is calculated.
Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
---
target/sh4/translate.c | 16 ++++++++++------
1 file changed, 10 insertions(+), 6 deletions(-)
diff --git a/target/sh4/translate.c b/target/sh4/translate.c
index 47c0f3404e..b0ac631859 100644
--- a/target/sh4/translate.c
+++ b/target/sh4/translate.c
@@ -933,16 +933,20 @@ static void _decode_opc(DisasContext * ctx)
return;
case 0x300b: /* subv Rm,Rn */
{
- TCGv t0, t1, t2;
- t0 = tcg_temp_new();
- tcg_gen_sub_i32(t0, REG(B11_8), REG(B7_4));
+ TCGv Rn = REG(B11_8);
+ TCGv Rm = REG(B7_4);
+ TCGv result, t1, t2;
+
+ result = tcg_temp_new();
t1 = tcg_temp_new();
- tcg_gen_xor_i32(t1, t0, REG(B11_8));
t2 = tcg_temp_new();
- tcg_gen_xor_i32(t2, REG(B11_8), REG(B7_4));
+ tcg_gen_add_i32(result, Rn, Rm);
+ /* T = ((Rn ^ Rm) & (Result ^ Rn)) >> 31 */
+ tcg_gen_xor_i32(t1, result, Rn);
+ tcg_gen_xor_i32(t2, Rn, Rm);
tcg_gen_and_i32(t1, t1, t2);
tcg_gen_shri_i32(cpu_sr_t, t1, 31);
- tcg_gen_mov_i32(REG(B11_8), t0);
+ tcg_gen_mov_i32(Rm, result);
}
return;
case 0x2008: /* tst Rm,Rn */
--
2.41.0
^ permalink raw reply related [flat|nested] 6+ messages in thread
* Re: [PATCH v3 0/4] target/sh4: Fix ADDV/SUBV opcodes
2024-04-30 14:56 [PATCH v3 0/4] target/sh4: Fix ADDV/SUBV opcodes Philippe Mathieu-Daudé
` (3 preceding siblings ...)
2024-04-30 14:56 ` [PATCH v3 4/4] target/sh4: Rename TCGv variables as manual for SUBV opcode Philippe Mathieu-Daudé
@ 2024-04-30 15:42 ` Richard Henderson
4 siblings, 0 replies; 6+ messages in thread
From: Richard Henderson @ 2024-04-30 15:42 UTC (permalink / raw)
To: Philippe Mathieu-Daudé, qemu-devel
Cc: Yoshinori Sato, Paul Cercueil, John Paul Adrian Glaubitz
On 4/30/24 07:56, Philippe Mathieu-Daudé wrote:
> Philippe Mathieu-Daudé (4):
> target/sh4: Fix ADDV opcode
> target/sh4: Fix SUBV opcode
> target/sh4: Rename TCGv variables as manual for ADDV opcode
> target/sh4: Rename TCGv variables as manual for SUBV opcode
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
r~
^ permalink raw reply [flat|nested] 6+ messages in thread
end of thread, other threads:[~2024-04-30 15:42 UTC | newest]
Thread overview: 6+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2024-04-30 14:56 [PATCH v3 0/4] target/sh4: Fix ADDV/SUBV opcodes Philippe Mathieu-Daudé
2024-04-30 14:56 ` [PATCH v3 1/4] target/sh4: Fix ADDV opcode Philippe Mathieu-Daudé
2024-04-30 14:56 ` [PATCH v3 2/4] target/sh4: Fix SUBV opcode Philippe Mathieu-Daudé
2024-04-30 14:56 ` [PATCH v3 3/4] target/sh4: Rename TCGv variables as manual for ADDV opcode Philippe Mathieu-Daudé
2024-04-30 14:56 ` [PATCH v3 4/4] target/sh4: Rename TCGv variables as manual for SUBV opcode Philippe Mathieu-Daudé
2024-04-30 15:42 ` [PATCH v3 0/4] target/sh4: Fix ADDV/SUBV opcodes Richard Henderson
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).