Add a vDSO function __vdso_riscv_hwprobe, which can sit in front of the
riscv_hwprobe syscall and answer common queries. We stash a copy of
static answers for the "all CPUs" case in the vDSO data page. This data
is private to the vDSO, so we can decide later to change what's stored
there or under what conditions we defer to the syscall. Currently all
data can be discovered at boot, so the vDSO function answers all queries
when the cpumask is set to the "all CPUs" hint.
There's also a boolean in the data that lets the vDSO function know that
all CPUs are the same. In that case, the vDSO will also answer queries
for arbitrary CPU masks in addition to the "all CPUs" hint.
Signed-off-by: Evan Green <evan@rivosinc.com>
---
Changes in v4:
- Added a prototype to fix -Wmissing-prototype warning (lkp@intel.com)
- Fixed rv32 build failure (lkp@intel.com)
- Make vdso prototype match syscall types update
Changes in v3:
- Introduced vDSO function
One aspect of this that's less than perfect is that there are two copies
of the arch_vdso_data in the data page. This stems from the fact that
vdso_data is arrayed by CS_BASES. Since we're very far away from filling
up the page, and this is not ABI, I left it for now. If things get full
we can move this data out to its own page, or refactor vdso_data
tree-wide to allow for non-arrayed data.
---
arch/riscv/Kconfig | 1 +
arch/riscv/include/asm/vdso/data.h | 17 +++++++
arch/riscv/include/asm/vdso/gettimeofday.h | 8 ++++
arch/riscv/kernel/sys_riscv.c | 41 +++++++++++++++++
arch/riscv/kernel/vdso.c | 6 ---
arch/riscv/kernel/vdso/Makefile | 2 +
arch/riscv/kernel/vdso/hwprobe.c | 52 ++++++++++++++++++++++
arch/riscv/kernel/vdso/sys_hwprobe.S | 15 +++++++
arch/riscv/kernel/vdso/vdso.lds.S | 1 +
9 files changed, 137 insertions(+), 6 deletions(-)
create mode 100644 arch/riscv/include/asm/vdso/data.h
create mode 100644 arch/riscv/kernel/vdso/hwprobe.c
create mode 100644 arch/riscv/kernel/vdso/sys_hwprobe.S
@@ -33,6 +33,7 @@ config RISCV
select ARCH_HAS_STRICT_MODULE_RWX if MMU && !XIP_KERNEL
select ARCH_HAS_TICK_BROADCAST if GENERIC_CLOCKEVENTS_BROADCAST
select ARCH_HAS_UBSAN_SANITIZE_ALL
+ select ARCH_HAS_VDSO_DATA
select ARCH_OPTIONAL_KERNEL_RWX if ARCH_HAS_STRICT_KERNEL_RWX
select ARCH_OPTIONAL_KERNEL_RWX_DEFAULT
select ARCH_STACKWALK
new file mode 100644
@@ -0,0 +1,17 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+#ifndef __RISCV_ASM_VDSO_DATA_H
+#define __RISCV_ASM_VDSO_DATA_H
+
+#include <linux/types.h>
+#include <vdso/datapage.h>
+#include <asm/hwprobe.h>
+
+struct arch_vdso_data {
+ /* Stash static answers to the hwprobe queries when all CPUs are selected. */
+ __u64 all_cpu_hwprobe_values[RISCV_HWPROBE_MAX_KEY + 1];
+
+ /* Boolean indicating all CPUs have the same static hwprobe values. */
+ __u8 homogeneous_cpus;
+};
+
+#endif /* __RISCV_ASM_VDSO_DATA_H */
@@ -9,6 +9,12 @@
#include <asm/csr.h>
#include <uapi/linux/time.h>
+/*
+ * 32-bit land is lacking generic time vsyscalls as well as the legacy 32-bit
+ * time syscalls like gettimeofday. Skip these definitions since on 32-bit.
+ */
+#ifdef CONFIG_GENERIC_TIME_VSYSCALL
+
#define VDSO_HAS_CLOCK_GETRES 1
static __always_inline
@@ -60,6 +66,8 @@ int clock_getres_fallback(clockid_t _clkid, struct __kernel_timespec *_ts)
return ret;
}
+#endif /* CONFIG_GENERIC_TIME_VSYSCALL */
+
static __always_inline u64 __arch_get_hw_counter(s32 clock_mode,
const struct vdso_data *vd)
{
@@ -14,6 +14,7 @@
#include <asm/uaccess.h>
#include <asm/unistd.h>
#include <asm-generic/mman-common.h>
+#include <vdso/vsyscall.h>
static long riscv_sys_mmap(unsigned long addr, unsigned long len,
unsigned long prot, unsigned long flags,
@@ -244,6 +245,46 @@ static int do_riscv_hwprobe(struct riscv_hwprobe __user *pairs,
}
+static int __init init_hwprobe_vdso_data(void)
+{
+ struct vdso_data *vd = __arch_get_k_vdso_data();
+ struct arch_vdso_data *avd = &vd->arch_data;
+ u64 id_bitsmash = 0;
+ struct riscv_hwprobe pair;
+ int key;
+
+ /*
+ * Initialize vDSO data with the answers for the "all CPUs" case, to
+ * save a syscall in the common case.
+ */
+ for (key = 0; key <= RISCV_HWPROBE_MAX_KEY; key++) {
+ pair.key = key;
+ hwprobe_one_pair(&pair, cpu_online_mask);
+
+ WARN_ON_ONCE(pair.key < 0);
+
+ avd->all_cpu_hwprobe_values[key] = pair.value;
+ /*
+ * Smash together the vendor, arch, and impl IDs to see if
+ * they're all 0 or any negative.
+ */
+ if (key <= RISCV_HWPROBE_KEY_MIMPID)
+ id_bitsmash |= pair.value;
+ }
+
+ /*
+ * If the arch, vendor, and implementation ID are all the same across
+ * all harts, then assume all CPUs are the same, and allow the vDSO to
+ * answer queries for arbitrary masks. However if all values are 0 (not
+ * populated) or any value returns -1 (varies across CPUs), then the
+ * vDSO should defer to the kernel for exotic cpu masks.
+ */
+ avd->homogeneous_cpus = (id_bitsmash > 0);
+ return 0;
+}
+
+arch_initcall_sync(init_hwprobe_vdso_data);
+
SYSCALL_DEFINE5(riscv_hwprobe, struct riscv_hwprobe __user *, pairs,
size_t, pair_count, size_t, cpu_count, unsigned long __user *,
cpus, unsigned int, flags)
@@ -14,13 +14,7 @@
#include <asm/page.h>
#include <asm/vdso.h>
#include <linux/time_namespace.h>
-
-#ifdef CONFIG_GENERIC_TIME_VSYSCALL
#include <vdso/datapage.h>
-#else
-struct vdso_data {
-};
-#endif
extern char vdso_start[], vdso_end[];
#ifdef CONFIG_COMPAT
@@ -12,6 +12,8 @@ vdso-syms += vgettimeofday
endif
vdso-syms += getcpu
vdso-syms += flush_icache
+vdso-syms += hwprobe
+vdso-syms += sys_hwprobe
# Files to link into the vdso
obj-vdso = $(patsubst %, %.o, $(vdso-syms)) note.o
new file mode 100644
@@ -0,0 +1,52 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * Copyright 2023 Rivos, Inc
+ */
+
+#include <linux/types.h>
+#include <vdso/datapage.h>
+#include <vdso/helpers.h>
+
+extern int riscv_hwprobe(struct riscv_hwprobe *pairs, size_t pair_count,
+ size_t cpu_count, unsigned long *cpus,
+ unsigned int flags);
+
+/* Add a prototype to avoid -Wmissing-prototypes warning. */
+int __vdso_riscv_hwprobe(struct riscv_hwprobe *pairs, size_t pair_count,
+ size_t cpu_count, unsigned long *cpus,
+ unsigned int flags);
+
+int __vdso_riscv_hwprobe(struct riscv_hwprobe *pairs, size_t pair_count,
+ size_t cpu_count, unsigned long *cpus,
+ unsigned int flags)
+{
+ const struct vdso_data *vd = __arch_get_vdso_data();
+ const struct arch_vdso_data *avd = &vd->arch_data;
+ bool all_cpus = !cpu_count && !cpus;
+ struct riscv_hwprobe *p = pairs;
+ struct riscv_hwprobe *end = pairs + pair_count;
+
+ /*
+ * Defer to the syscall for exotic requests. The vdso has answers
+ * stashed away only for the "all cpus" case. If all CPUs are
+ * homogeneous, then this function can handle requests for arbitrary
+ * masks.
+ */
+ if ((flags != 0) || (!all_cpus && !avd->homogeneous_cpus))
+ return riscv_hwprobe(pairs, pair_count, cpu_count, cpus, flags);
+
+ /* This is something we can handle, fill out the pairs. */
+ while (p < end) {
+ if (p->key <= RISCV_HWPROBE_MAX_KEY) {
+ p->value = avd->all_cpu_hwprobe_values[p->key];
+
+ } else {
+ p->key = -1;
+ p->value = 0;
+ }
+
+ p++;
+ }
+
+ return 0;
+}
new file mode 100644
@@ -0,0 +1,15 @@
+/* SPDX-License-Identifier: GPL-2.0 */
+/* Copyright (C) 2022 Rivos, Inc */
+
+#include <linux/linkage.h>
+#include <asm/unistd.h>
+
+.text
+ENTRY(riscv_hwprobe)
+ .cfi_startproc
+ li a7, __NR_riscv_hwprobe
+ ecall
+ ret
+
+ .cfi_endproc
+ENDPROC(riscv_hwprobe)
@@ -75,6 +75,7 @@ VERSION
#endif
__vdso_getcpu;
__vdso_flush_icache;
+ __vdso_riscv_hwprobe;
local: *;
};
}