[v3,6/7] riscv: add a data fence for CMODX in the kernel mode

Commit Message

Andy Chiu Nov. 27, 2024, 5:29 p.m. UTC

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.

Thus, add a fence here to order data writes before making the IPI.

Signed-off-by: Andy Chiu <[e-mail]>
---
 arch/riscv/mm/cacheflush.c | 15 ++++++++++++++-
 1 file changed, 14 insertions(+), 1 deletion(-)

Patch

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);