From: Peter Maydell <peter.maydell@linaro.org>
To: qemu-devel@nongnu.org
Subject: [Qemu-devel] [PULL 35/37] target/arm: Emit barriers for A32/T32 load-acquire/store-release insns
Date: Mon, 7 Jan 2019 16:31:15 +0000 [thread overview]
Message-ID: <20190107163117.16269-36-peter.maydell@linaro.org> (raw)
In-Reply-To: <20190107163117.16269-1-peter.maydell@linaro.org>
Now that MTTCG is here, the comment in the 32-bit Arm decoder that
"Since the emulation does not have barriers, the acquire/release
semantics need no special handling" is no longer true. Emit the
correct barriers for the load-acquire/store-release insns, as
we already do in the A64 decoder.
Signed-off-by: Peter Maydell <peter.maydell@linaro.org>
Tested-by: Alex Bennée <alex.bennee@linaro.org>
Reviewed-by: Alex Bennée <alex.bennee@linaro.org>
---
target/arm/translate.c | 33 ++++++++++++++++++++++++++-------
1 file changed, 26 insertions(+), 7 deletions(-)
diff --git a/target/arm/translate.c b/target/arm/translate.c
index ed3db0c3946..66cf28c8cbe 100644
--- a/target/arm/translate.c
+++ b/target/arm/translate.c
@@ -9733,6 +9733,8 @@ static void disas_arm_insn(DisasContext *s, unsigned int insn)
rd = (insn >> 12) & 0xf;
if (insn & (1 << 23)) {
/* load/store exclusive */
+ bool is_ld = extract32(insn, 20, 1);
+ bool is_lasr = !extract32(insn, 8, 1);
int op2 = (insn >> 8) & 3;
op1 = (insn >> 21) & 0x3;
@@ -9760,11 +9762,12 @@ static void disas_arm_insn(DisasContext *s, unsigned int insn)
addr = tcg_temp_local_new_i32();
load_reg_var(s, addr, rn);
- /* Since the emulation does not have barriers,
- the acquire/release semantics need no special
- handling */
+ if (is_lasr && !is_ld) {
+ tcg_gen_mb(TCG_MO_ALL | TCG_BAR_STRL);
+ }
+
if (op2 == 0) {
- if (insn & (1 << 20)) {
+ if (is_ld) {
tmp = tcg_temp_new_i32();
switch (op1) {
case 0: /* lda */
@@ -9810,7 +9813,7 @@ static void disas_arm_insn(DisasContext *s, unsigned int insn)
}
tcg_temp_free_i32(tmp);
}
- } else if (insn & (1 << 20)) {
+ } else if (is_ld) {
switch (op1) {
case 0: /* ldrex */
gen_load_exclusive(s, rd, 15, addr, 2);
@@ -9847,6 +9850,10 @@ static void disas_arm_insn(DisasContext *s, unsigned int insn)
}
}
tcg_temp_free_i32(addr);
+
+ if (is_lasr && is_ld) {
+ tcg_gen_mb(TCG_MO_ALL | TCG_BAR_LDAQ);
+ }
} else if ((insn & 0x00300f00) == 0) {
/* 0bcccc_0001_0x00_xxxx_xxxx_0000_1001_xxxx
* - SWP, SWPB
@@ -10862,6 +10869,8 @@ static void disas_thumb2_insn(DisasContext *s, uint32_t insn)
tcg_gen_addi_i32(tmp, tmp, s->pc);
store_reg(s, 15, tmp);
} else {
+ bool is_lasr = false;
+ bool is_ld = extract32(insn, 20, 1);
int op2 = (insn >> 6) & 0x3;
op = (insn >> 4) & 0x3;
switch (op2) {
@@ -10883,12 +10892,18 @@ static void disas_thumb2_insn(DisasContext *s, uint32_t insn)
case 3:
/* Load-acquire/store-release exclusive */
ARCH(8);
+ is_lasr = true;
break;
}
+
+ if (is_lasr && !is_ld) {
+ tcg_gen_mb(TCG_MO_ALL | TCG_BAR_STRL);
+ }
+
addr = tcg_temp_local_new_i32();
load_reg_var(s, addr, rn);
if (!(op2 & 1)) {
- if (insn & (1 << 20)) {
+ if (is_ld) {
tmp = tcg_temp_new_i32();
switch (op) {
case 0: /* ldab */
@@ -10927,12 +10942,16 @@ static void disas_thumb2_insn(DisasContext *s, uint32_t insn)
}
tcg_temp_free_i32(tmp);
}
- } else if (insn & (1 << 20)) {
+ } else if (is_ld) {
gen_load_exclusive(s, rs, rd, addr, op);
} else {
gen_store_exclusive(s, rm, rs, rd, addr, op);
}
tcg_temp_free_i32(addr);
+
+ if (is_lasr && is_ld) {
+ tcg_gen_mb(TCG_MO_ALL | TCG_BAR_LDAQ);
+ }
}
} else {
/* Load/store multiple, RFE, SRS. */
--
2.19.2
next prev parent reply other threads:[~2019-01-07 16:32 UTC|newest]
Thread overview: 46+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-01-07 16:30 [Qemu-devel] [PULL 00/37] target-arm queue Peter Maydell
2019-01-07 16:30 ` [Qemu-devel] [PULL 01/37] target/arm: Convert ARM_TBFLAG_* to FIELDs Peter Maydell
2019-01-07 16:30 ` [Qemu-devel] [PULL 02/37] target/arm: SVE brk[ab] merging does not have s bit Peter Maydell
2019-01-07 16:30 ` [Qemu-devel] [PULL 03/37] hw/cpu: introduce CPU clusters Peter Maydell
2019-01-07 16:30 ` [Qemu-devel] [PULL 04/37] gdbstub: introduce GDB processes Peter Maydell
2019-01-07 16:30 ` [Qemu-devel] [PULL 05/37] gdbstub: add multiprocess support to '?' packets Peter Maydell
2019-01-07 16:30 ` [Qemu-devel] [PULL 06/37] gdbstub: add multiprocess support to 'H' and 'T' packets Peter Maydell
2019-01-17 18:13 ` Peter Maydell
2019-01-17 18:19 ` Peter Maydell
2019-01-07 16:30 ` [Qemu-devel] [PULL 07/37] gdbstub: add multiprocess support to vCont packets Peter Maydell
2019-01-07 16:30 ` [Qemu-devel] [PULL 08/37] gdbstub: add multiprocess support to 'sC' packets Peter Maydell
2019-01-07 16:30 ` [Qemu-devel] [PULL 09/37] gdbstub: add multiprocess support to (f|s)ThreadInfo and ThreadExtraInfo Peter Maydell
2019-01-07 16:30 ` [Qemu-devel] [PULL 10/37] gdbstub: add multiprocess support to Xfer:features:read: Peter Maydell
2019-01-17 17:22 ` Peter Maydell
2019-01-07 16:30 ` [Qemu-devel] [PULL 11/37] gdbstub: add multiprocess support to gdb_vm_state_change() Peter Maydell
2019-01-07 16:30 ` [Qemu-devel] [PULL 12/37] gdbstub: add multiprocess support to 'D' packets Peter Maydell
2019-01-07 16:30 ` [Qemu-devel] [PULL 13/37] gdbstub: add support for extended mode packet Peter Maydell
2019-01-07 16:30 ` [Qemu-devel] [PULL 14/37] gdbstub: add support for vAttach packets Peter Maydell
2019-01-07 16:30 ` [Qemu-devel] [PULL 15/37] gdbstub: processes initialization on new peer connection Peter Maydell
2019-01-07 16:30 ` [Qemu-devel] [PULL 16/37] gdbstub: gdb_set_stop_cpu: ignore request when process is not attached Peter Maydell
2019-01-07 16:30 ` [Qemu-devel] [PULL 17/37] gdbstub: add multiprocess extension support Peter Maydell
2019-01-07 16:30 ` [Qemu-devel] [PULL 18/37] arm/xlnx-zynqmp: put APUs and RPUs in separate CPU clusters Peter Maydell
2019-01-07 16:30 ` [Qemu-devel] [PULL 19/37] Revert "armv7m: Guard against no -kernel argument" Peter Maydell
2019-01-07 16:31 ` [Qemu-devel] [PULL 20/37] hw/arm: versal: Plug memory leaks Peter Maydell
2019-01-07 16:31 ` [Qemu-devel] [PULL 21/37] MAINTAINERS: Add ARM-related files for hw/[misc|input|timer]/ Peter Maydell
2019-01-07 16:31 ` [Qemu-devel] [PULL 22/37] cpus.c: Fix race condition in cpu_stop_current() Peter Maydell
2019-01-07 16:31 ` [Qemu-devel] [PULL 23/37] hw/arm/allwinner-a10: Add the 'A' SRAM and the SRAM controller Peter Maydell
2019-01-07 16:31 ` [Qemu-devel] [PULL 24/37] qtest: Add set_irq_in command to set IRQ/GPIO level Peter Maydell
2019-01-09 5:58 ` Matthew Ogilvie
2019-01-10 15:53 ` Peter Maydell
2019-01-10 16:26 ` Eric Blake
2019-01-07 16:31 ` [Qemu-devel] [PULL 25/37] arm: Add header to host common definition for nRF51 SOC peripherals Peter Maydell
2019-01-07 16:31 ` [Qemu-devel] [PULL 26/37] hw/misc/nrf51_rng: Add NRF51 random number generator peripheral Peter Maydell
2019-01-07 16:31 ` [Qemu-devel] [PULL 27/37] arm: Instantiate NRF51 random number generator Peter Maydell
2019-01-07 16:31 ` [Qemu-devel] [PULL 28/37] hw/gpio/nrf51_gpio: Add nRF51 GPIO peripheral Peter Maydell
2019-01-07 16:31 ` [Qemu-devel] [PULL 29/37] arm: Instantiate NRF51 general purpose I/O Peter Maydell
2019-01-07 16:31 ` [Qemu-devel] [PULL 30/37] tests/microbit-test: Add Tests for nRF51 GPIO Peter Maydell
2019-01-07 16:31 ` [Qemu-devel] [PULL 31/37] hw/timer/nrf51_timer: Add nRF51 Timer peripheral Peter Maydell
2019-01-07 16:31 ` [Qemu-devel] [PULL 32/37] arm: Instantiate NRF51 Timers Peter Maydell
2019-01-07 16:31 ` [Qemu-devel] [PULL 33/37] tests/microbit-test: Add Tests for nRF51 Timer Peter Maydell
2019-01-07 16:31 ` [Qemu-devel] [PULL 34/37] arm: Add Clock peripheral stub to NRF51 SOC Peter Maydell
2019-01-07 16:31 ` Peter Maydell [this message]
2019-01-07 16:31 ` [Qemu-devel] [PULL 36/37] hw/misc/tz-mpc: Fix value of BLK_MAX register Peter Maydell
2019-01-07 16:31 ` [Qemu-devel] [PULL 37/37] Support u-boot noload images for arm as used by, NetBSD/evbarm GENERIC kernel Peter Maydell
2019-01-07 18:24 ` [Qemu-devel] [PULL 00/37] target-arm queue Peter Maydell
2019-01-07 20:29 ` no-reply
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20190107163117.16269-36-peter.maydell@linaro.org \
--to=peter.maydell@linaro.org \
--cc=qemu-devel@nongnu.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).