ftrace: Synchronize the initialization of ftrace_opsgrafted

There's some internal state that ftrace_ops needs to have set, but since
it can be declared outside of the ftrace.c code, it calls
ftrace_ops_init() on the ops in every global function. The issue is that
if two tasks call it on the same ops at the same time it is possible to
have the initialization of one corrupt the initialization of the other
call.

Create a ops_mutex to use to synchronize every initialization of the
ftrace_ops. The mutex is taken within checking the ftrace_ops flag that
states it was initializied but the flag is checked again after the mutex
has been taken. Checking first outside the mutex allows it to shortcut
having to take the mutex. But then the check needs to be done again after
the mute is taken in case of races.

Cc: stable@vger.kernel.org
Link: https://patch.msgid.link/20260902095501.6b59af20@gandalf.local.home
Fixes: f04f24fb7e48d ("ftrace, kprobes: Fix a deadlock on ftrace_regex_lock")
Reported-by: sashiko-bot@kernel.org
Close: https://lore.kernel.org/all/20260829025528.49A831F000E9@smtp.kernel.org/
Signed-off-by: Steven Rostedt <rostedt@goodmis.org>

1 files changed, 67 insertions, 0 deletions

diff --git a/lib/crypto/powerpc/sha1.h b/lib/crypto/powerpc/sha1.h
new file mode 100644
index 000000000..e2c010f03
--- /dev/null
+++ b/lib/crypto/powerpc/sha1.h
@@ -0,0 +1,67 @@
+/* SPDX-License-Identifier: GPL-2.0-or-later */
+/*
+ * SHA-1 optimized for PowerPC
+ *
+ * Copyright (c) 2015 Markus Stockhausen <stockhausen@collogia.de>
+ */
+
+#include <asm/switch_to.h>
+#include <linux/preempt.h>
+
+#ifdef CONFIG_SPE
+/*
+ * MAX_BYTES defines the number of bytes that are allowed to be processed
+ * between preempt_disable() and preempt_enable(). SHA1 takes ~1000
+ * operations per 64 bytes. e500 cores can issue two arithmetic instructions
+ * per clock cycle using one 32/64 bit unit (SU1) and one 32 bit unit (SU2).
+ * Thus 2KB of input data will need an estimated maximum of 18,000 cycles.
+ * Headroom for cache misses included. Even with the low end model clocked
+ * at 667 MHz this equals to a critical time window of less than 27us.
+ *
+ */
+#define MAX_BYTES 2048
+
+asmlinkage void ppc_spe_sha1_transform(struct sha1_block_state *state,
+				       const u8 *data, u32 nblocks);
+
+static void spe_begin(void)
+{
+	/* We just start SPE operations and will save SPE registers later. */
+	preempt_disable();
+	enable_kernel_spe();
+}
+
+static void spe_end(void)
+{
+	disable_kernel_spe();
+	/* reenable preemption */
+	preempt_enable();
+}
+
+static void sha1_blocks(struct sha1_block_state *state,
+			const u8 *data, size_t nblocks)
+{
+	do {
+		u32 unit = min_t(size_t, nblocks, MAX_BYTES / SHA1_BLOCK_SIZE);
+
+		spe_begin();
+		ppc_spe_sha1_transform(state, data, unit);
+		spe_end();
+
+		data += unit * SHA1_BLOCK_SIZE;
+		nblocks -= unit;
+	} while (nblocks);
+}
+#else /* CONFIG_SPE */
+asmlinkage void powerpc_sha_transform(struct sha1_block_state *state,
+				      const u8 data[SHA1_BLOCK_SIZE]);
+
+static void sha1_blocks(struct sha1_block_state *state,
+			const u8 *data, size_t nblocks)
+{
+	do {
+		powerpc_sha_transform(state, data);
+		data += SHA1_BLOCK_SIZE;
+	} while (--nblocks);
+}
+#endif /* !CONFIG_SPE */