riscv: Add support for userspace pointer masking

RISC-V supports pointer masking with a variable number of tag bits
(which is called "PMLEN" in the specification) and which is configured
at the next higher privilege level.

Wire up the PR_SET_TAGGED_ADDR_CTRL and PR_GET_TAGGED_ADDR_CTRL prctls
so userspace can request a lower bound on the number of tag bits and
determine the actual number of tag bits. As with arm64's
PR_TAGGED_ADDR_ENABLE, the pointer masking configuration is
thread-scoped, inherited on clone() and fork() and cleared on execve().

Reviewed-by: Charlie Jenkins <charlie@rivosinc.com>
Tested-by: Charlie Jenkins <charlie@rivosinc.com>
Signed-off-by: Samuel Holland <samuel.holland@sifive.com>
Link: https://lore.kernel.org/r/20241016202814.4061541-5-samuel.holland@sifive.com
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>

Author: SiFiveHolland

Documentation/arch/riscv/uabi.rst

@@ -68,3 +68,15 @@ Misaligned accesses
 Misaligned scalar accesses are supported in userspace, but they may perform
 poorly.  Misaligned vector accesses are only supported if the Zicclsm extension
 is supported.
+
+Pointer masking
+---------------
+
+Support for pointer masking in userspace (the Supm extension) is provided via
+the ``PR_SET_TAGGED_ADDR_CTRL`` and ``PR_GET_TAGGED_ADDR_CTRL`` ``prctl()``
+operations. Pointer masking is disabled by default. To enable it, userspace
+must call ``PR_SET_TAGGED_ADDR_CTRL`` with the ``PR_PMLEN`` field set to the
+number of mask/tag bits needed by the application. ``PR_PMLEN`` is interpreted
+as a lower bound; if the kernel is unable to satisfy the request, the
+``PR_SET_TAGGED_ADDR_CTRL`` operation will fail. The actual number of tag bits
+is returned in ``PR_PMLEN`` by the ``PR_GET_TAGGED_ADDR_CTRL`` operation.

arch/riscv/Kconfig

@@ -531,6 +531,17 @@ config RISCV_ISA_C
 
 	  If you don't know what to do here, say Y.
 
+config RISCV_ISA_SUPM
+	bool "Supm extension for userspace pointer masking"
+	depends on 64BIT
+	default y
+	help
+	  Add support for pointer masking in userspace (Supm) when the
+	  underlying hardware extension (Smnpm or Ssnpm) is detected at boot.
+
+	  If this option is disabled, userspace will be unable to use
+	  the prctl(PR_{SET,GET}_TAGGED_ADDR_CTRL) API.
+
 config RISCV_ISA_SVNAPOT
 	bool "Svnapot extension support for supervisor mode NAPOT pages"
 	depends on 64BIT && MMU

arch/riscv/include/asm/processor.h

@@ -178,6 +178,14 @@ extern int set_unalign_ctl(struct task_struct *tsk, unsigned int val);
 #define RISCV_SET_ICACHE_FLUSH_CTX(arg1, arg2)	riscv_set_icache_flush_ctx(arg1, arg2)
 extern int riscv_set_icache_flush_ctx(unsigned long ctx, unsigned long per_thread);
 
+#ifdef CONFIG_RISCV_ISA_SUPM
+/* PR_{SET,GET}_TAGGED_ADDR_CTRL prctl */
+long set_tagged_addr_ctrl(struct task_struct *task, unsigned long arg);
+long get_tagged_addr_ctrl(struct task_struct *task);
+#define SET_TAGGED_ADDR_CTRL(arg)	set_tagged_addr_ctrl(current, arg)
+#define GET_TAGGED_ADDR_CTRL()		get_tagged_addr_ctrl(current)
+#endif
+
 #endif /* __ASSEMBLY__ */
 
 #endif /* _ASM_RISCV_PROCESSOR_H */

arch/riscv/include/asm/switch_to.h

@@ -70,6 +70,17 @@ static __always_inline bool has_fpu(void) { return false; }
 #define __switch_to_fpu(__prev, __next) do { } while (0)
 #endif
 
+static inline void envcfg_update_bits(struct task_struct *task,
+				      unsigned long mask, unsigned long val)
+{
+	unsigned long envcfg;
+
+	envcfg = (task->thread.envcfg & ~mask) | val;
+	task->thread.envcfg = envcfg;
+	if (task == current)
+		csr_write(CSR_ENVCFG, envcfg);
+}
+
 static inline void __switch_to_envcfg(struct task_struct *next)
 {
 	asm volatile (ALTERNATIVE("nop", "csrw " __stringify(CSR_ENVCFG) ", %0",

arch/riscv/kernel/process.c

@@ -7,6 +7,7 @@
  * Copyright (C) 2017 SiFive
  */
 
+#include <linux/bitfield.h>
 #include <linux/cpu.h>
 #include <linux/kernel.h>
 #include <linux/sched.h>
@@ -180,6 +181,10 @@ void flush_thread(void)
 	memset(&current->thread.vstate, 0, sizeof(struct __riscv_v_ext_state));
 	clear_tsk_thread_flag(current, TIF_RISCV_V_DEFER_RESTORE);
 #endif
+#ifdef CONFIG_RISCV_ISA_SUPM
+	if (riscv_has_extension_unlikely(RISCV_ISA_EXT_SUPM))
+		envcfg_update_bits(current, ENVCFG_PMM, ENVCFG_PMM_PMLEN_0);
+#endif
 }
 
 void arch_release_task_struct(struct task_struct *tsk)
@@ -242,3 +247,89 @@ void __init arch_task_cache_init(void)
 {
 	riscv_v_setup_ctx_cache();
 }
+
+#ifdef CONFIG_RISCV_ISA_SUPM
+enum {
+	PMLEN_0 = 0,
+	PMLEN_7 = 7,
+	PMLEN_16 = 16,
+};
+
+static bool have_user_pmlen_7;
+static bool have_user_pmlen_16;
+
+long set_tagged_addr_ctrl(struct task_struct *task, unsigned long arg)
+{
+	unsigned long valid_mask = PR_PMLEN_MASK;
+	struct thread_info *ti = task_thread_info(task);
+	unsigned long pmm;
+	u8 pmlen;
+
+	if (is_compat_thread(ti))
+		return -EINVAL;
+
+	if (arg & ~valid_mask)
+		return -EINVAL;
+
+	/*
+	 * Prefer the smallest PMLEN that satisfies the user's request,
+	 * in case choosing a larger PMLEN has a performance impact.
+	 */
+	pmlen = FIELD_GET(PR_PMLEN_MASK, arg);
+	if (pmlen == PMLEN_0)
+		pmm = ENVCFG_PMM_PMLEN_0;
+	else if (pmlen <= PMLEN_7 && have_user_pmlen_7)
+		pmm = ENVCFG_PMM_PMLEN_7;
+	else if (pmlen <= PMLEN_16 && have_user_pmlen_16)
+		pmm = ENVCFG_PMM_PMLEN_16;
+	else
+		return -EINVAL;
+
+	envcfg_update_bits(task, ENVCFG_PMM, pmm);
+
+	return 0;
+}
+
+long get_tagged_addr_ctrl(struct task_struct *task)
+{
+	struct thread_info *ti = task_thread_info(task);
+	long ret = 0;
+
+	if (is_compat_thread(ti))
+		return -EINVAL;
+
+	switch (task->thread.envcfg & ENVCFG_PMM) {
+	case ENVCFG_PMM_PMLEN_7:
+		ret = FIELD_PREP(PR_PMLEN_MASK, PMLEN_7);
+		break;
+	case ENVCFG_PMM_PMLEN_16:
+		ret = FIELD_PREP(PR_PMLEN_MASK, PMLEN_16);
+		break;
+	}
+
+	return ret;
+}
+
+static bool try_to_set_pmm(unsigned long value)
+{
+	csr_set(CSR_ENVCFG, value);
+	return (csr_read_clear(CSR_ENVCFG, ENVCFG_PMM) & ENVCFG_PMM) == value;
+}
+
+static int __init tagged_addr_init(void)
+{
+	if (!riscv_has_extension_unlikely(RISCV_ISA_EXT_SUPM))
+		return 0;
+
+	/*
+	 * envcfg.PMM is a WARL field. Detect which values are supported.
+	 * Assume the supported PMLEN values are the same on all harts.
+	 */
+	csr_clear(CSR_ENVCFG, ENVCFG_PMM);
+	have_user_pmlen_7 = try_to_set_pmm(ENVCFG_PMM_PMLEN_7);
+	have_user_pmlen_16 = try_to_set_pmm(ENVCFG_PMM_PMLEN_16);
+
+	return 0;
+}
+core_initcall(tagged_addr_init);
+#endif	/* CONFIG_RISCV_ISA_SUPM */

include/uapi/linux/prctl.h

@@ -230,7 +230,7 @@ struct prctl_mm_map {
 # define PR_PAC_APDBKEY			(1UL << 3)
 # define PR_PAC_APGAKEY			(1UL << 4)
 
-/* Tagged user address controls for arm64 */
+/* Tagged user address controls for arm64 and RISC-V */
 #define PR_SET_TAGGED_ADDR_CTRL		55
 #define PR_GET_TAGGED_ADDR_CTRL		56
 # define PR_TAGGED_ADDR_ENABLE		(1UL << 0)
@@ -244,6 +244,9 @@ struct prctl_mm_map {
 # define PR_MTE_TAG_MASK		(0xffffUL << PR_MTE_TAG_SHIFT)
 /* Unused; kept only for source compatibility */
 # define PR_MTE_TCF_SHIFT		1
+/* RISC-V pointer masking tag length */
+# define PR_PMLEN_SHIFT			24
+# define PR_PMLEN_MASK			(0x7fUL << PR_PMLEN_SHIFT)
 
 /* Control reclaim behavior when allocating memory */
 #define PR_SET_IO_FLUSHER		57