From: Andy Chiu <andybnac@gmail.com>
To: linux-riscv@lists.infradead.org, alexghiti@rivosinc.com,
palmer@dabbelt.com
Cc: "Andy Chiu" <andybnac@gmail.com>,
"Björn Töpel" <bjorn@rivosinc.com>,
linux-kernel@vger.kernel.org, "Alexandre Ghiti" <alex@ghiti.fr>,
puranjay12@gmail.com, paul.walmsley@sifive.com,
greentime.hu@sifive.com, nick.hu@sifive.com,
nylon.chen@sifive.com, eric.lin@sifive.com,
vicent.chen@sifive.com, zong.li@sifive.com,
yongxuan.wang@sifive.com, samuel.holland@sifive.com,
olivia.chu@sifive.com, c2232430@gmail.com
Subject: [PATCH v4 08/12] riscv: add a data fence for CMODX in the kernel mode
Date: Tue, 8 Apr 2025 02:08:32 +0800 [thread overview]
Message-ID: <20250407180838.42877-8-andybnac@gmail.com> (raw)
In-Reply-To: <20250407180838.42877-1-andybnac@gmail.com>
RISC-V spec explicitly calls out that a local fence.i is not enough for
the code modification to be visble from a remote hart. In fact, it
states:
To make a store to instruction memory visible to all RISC-V harts, the
writing hart also has to execute a data FENCE before requesting that all
remote RISC-V harts execute a FENCE.I.
Although current riscv drivers for IPI use ordered MMIO when sending IPIs
in order to synchronize the action between previous csd writes, riscv
does not restrict itself to any particular flavor of IPI. Any driver or
firmware implementation that does not order data writes before the IPI
may pose a risk for code-modifying race.
Thus, add a fence here to order data writes before making the IPI.
Signed-off-by: Andy Chiu <andybnac@gmail.com>
Reviewed-by: Björn Töpel <bjorn@rivosinc.com>
---
Changelog v4:
- Explain more in commit msg and add Björn's R-b
---
arch/riscv/mm/cacheflush.c | 15 ++++++++++++++-
1 file changed, 14 insertions(+), 1 deletion(-)
diff --git a/arch/riscv/mm/cacheflush.c b/arch/riscv/mm/cacheflush.c
index b81672729887..b2e4b81763f8 100644
--- a/arch/riscv/mm/cacheflush.c
+++ b/arch/riscv/mm/cacheflush.c
@@ -24,7 +24,20 @@ void flush_icache_all(void)
if (num_online_cpus() < 2)
return;
- else if (riscv_use_sbi_for_rfence())
+
+ /*
+ * Make sure all previous writes to the D$ are ordered before making
+ * the IPI. The RISC-V spec states that a hart must execute a data fence
+ * before triggering a remote fence.i in order to make the modification
+ * visable for remote harts.
+ *
+ * IPIs on RISC-V are triggered by MMIO writes to either CLINT or
+ * S-IMSIC, so the fence ensures previous data writes "happen before"
+ * the MMIO.
+ */
+ RISCV_FENCE(w, o);
+
+ if (riscv_use_sbi_for_rfence())
sbi_remote_fence_i(NULL);
else
on_each_cpu(ipi_remote_fence_i, NULL, 1);
--
2.39.3 (Apple Git-145)
_______________________________________________
linux-riscv mailing list
linux-riscv@lists.infradead.org
http://lists.infradead.org/mailman/listinfo/linux-riscv
next prev parent reply other threads:[~2025-04-07 18:52 UTC|newest]
Thread overview: 30+ messages / expand[flat|nested] mbox.gz Atom feed top
2025-04-07 18:08 [PATCH v4 01/12] riscv: ftrace: support fastcc in Clang for WITH_ARGS Andy Chiu
2025-04-07 18:08 ` [PATCH v4 02/12] riscv: ftrace factor out code defined by !WITH_ARG Andy Chiu
2025-04-07 18:08 ` [PATCH v4 03/12] riscv: ftrace: align patchable functions to 4 Byte boundary Andy Chiu
2025-04-07 18:08 ` [PATCH v4 04/12] kernel: ftrace: export ftrace_sync_ipi Andy Chiu
2025-04-08 22:31 ` kernel test robot
2025-04-23 8:13 ` Alexandre Ghiti
2025-04-07 18:08 ` [PATCH v4 05/12] riscv: ftrace: prepare ftrace for atomic code patching Andy Chiu
2025-04-11 13:15 ` Robbin Ehn
2025-04-23 8:22 ` Alexandre Ghiti
2025-05-05 14:06 ` Alexandre Ghiti
2025-05-07 14:18 ` Andy Chiu
2025-05-07 14:35 ` Alexandre Ghiti
2025-04-07 18:08 ` [PATCH v4 06/12] riscv: ftrace: do not use stop_machine to update code Andy Chiu
2025-04-07 18:08 ` [PATCH v4 07/12] riscv: vector: Support calling schedule() for preemptible Vector Andy Chiu
2025-04-07 18:08 ` Andy Chiu [this message]
2025-04-07 18:08 ` [PATCH v4 09/12] riscv: ftrace: support PREEMPT Andy Chiu
2025-04-07 18:08 ` [PATCH v4 10/12] riscv: Implement HAVE_DYNAMIC_FTRACE_WITH_CALL_OPS Andy Chiu
2026-02-21 12:15 ` Conor Dooley
2026-02-23 15:18 ` Puranjay Mohan
2026-02-23 15:27 ` Conor Dooley
2026-02-23 15:41 ` Puranjay Mohan
2026-02-23 16:29 ` Conor Dooley
2026-02-23 17:36 ` Puranjay Mohan
2026-02-23 17:41 ` Conor Dooley
2025-04-07 18:08 ` [PATCH v4 11/12] riscv: ftrace: support direct call using call_ops Andy Chiu
2025-04-07 18:08 ` [PATCH v4 12/12] riscv: Documentation: add a description about dynamic ftrace Andy Chiu
2025-04-11 12:02 ` Robbin Ehn
2025-04-10 20:05 ` [PATCH v4 01/12] riscv: ftrace: support fastcc in Clang for WITH_ARGS Björn Töpel
2025-05-07 13:58 ` Andy Chiu
2025-06-02 22:12 ` patchwork-bot+linux-riscv
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=20250407180838.42877-8-andybnac@gmail.com \
--to=andybnac@gmail.com \
--cc=alex@ghiti.fr \
--cc=alexghiti@rivosinc.com \
--cc=bjorn@rivosinc.com \
--cc=c2232430@gmail.com \
--cc=eric.lin@sifive.com \
--cc=greentime.hu@sifive.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-riscv@lists.infradead.org \
--cc=nick.hu@sifive.com \
--cc=nylon.chen@sifive.com \
--cc=olivia.chu@sifive.com \
--cc=palmer@dabbelt.com \
--cc=paul.walmsley@sifive.com \
--cc=puranjay12@gmail.com \
--cc=samuel.holland@sifive.com \
--cc=vicent.chen@sifive.com \
--cc=yongxuan.wang@sifive.com \
--cc=zong.li@sifive.com \
/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