linux-kernel - [PATCH v3 5/6] uapi: futex: Add a futex waitv syscall

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]

Message-Id: <20211126060024.3290177-5-alistair.francis@opensource.wdc.com>
Date:   Fri, 26 Nov 2021 16:00:23 +1000
From:   Alistair Francis <alistair.francis@...nsource.wdc.com>
To:     linux-kernel@...r.kernel.org
Cc:     tglx@...utronix.de, arnd@...db.de, namhyung@...nel.org,
        peterz@...radead.org, alistair23@...il.com, jolsa@...hat.com,
        dave@...olabs.net, mingo@...hat.com, dvhart@...radead.org,
        acme@...nel.org, linux-perf-users@...r.kernel.org,
        mark.rutland@....com, alexander.shishkin@...ux.intel.com,
        Alistair Francis <alistair.francis@....com>
Subject: [PATCH v3 5/6] uapi: futex: Add a futex waitv syscall

From: Alistair Francis <alistair.francis@....com>

This commit adds a futex waitv syscall wrapper that is exposed to
userspace.

Neither the kernel or glibc currently expose a futex wrapper, so
userspace is left performing raw syscalls. As the futex_waitv syscall
always expects a 64-bit time_t this can be tricky for 32-bit systems to
get correct.

In order to avoid userspace incorrectly passing the wrong timeouts let's
expose a public helper function that ensures the kernel is passed the
correct timeout struct.

Signed-off-by: Alistair Francis <alistair.francis@....com>
---
 include/uapi/linux/futex_syscall.h | 32 +++++++++++++++++++++++++++++-
 1 file changed, 31 insertions(+), 1 deletion(-)

diff --git a/include/uapi/linux/futex_syscall.h b/include/uapi/linux/futex_syscall.h
index e0305c581142..98c550990833 100644
--- a/include/uapi/linux/futex_syscall.h
+++ b/include/uapi/linux/futex_syscall.h
@@ -9,12 +9,15 @@
 #ifndef _UAPI_LINUX_FUTEX_SYSCALL_H
 #define _UAPI_LINUX_FUTEX_SYSCALL_H
 
-#include <asm/unistd.h>
+#include <unistd.h>
 #include <errno.h>
+#include <linux/futex.h>
 #include <linux/types.h>
 #include <linux/time_types.h>
 #include <stdint.h>
 #include <sys/syscall.h>
+#include <sys/types.h>
+#include <time.h>
 
 /**
  * futex_syscall_timeout() - __NR_futex/__NR_futex_time64 syscall wrapper
@@ -86,4 +89,31 @@ __kernel_futex_syscall_nr_requeue(volatile uint32_t *uaddr, int op, uint32_t val
 	return -1;
 }
 
+/**
+ * futex_waitv - Wait at multiple futexes, wake on any
+ * @waiters:    Array of waiters
+ * @nr_waiters: Length of waiters array
+ * @flags: Operation flags
+ * @timo:  Optional timeout for operation
+ */
+static inline int
+__kernel_futex_syscall_waitv(volatile struct futex_waitv *waiters, unsigned long nr_waiters,
+			      unsigned long flags, struct timespec *timo, clockid_t clockid)
+{
+	/* futex_waitv expects a 64-bit time_t */
+	if (sizeof(*timo) == sizeof(struct __kernel_timespec))
+		return syscall(__NR_futex_waitv, waiters, nr_waiters, flags, timo, clockid);
+
+	/* If the caller supplied a 32-bit time_t, convert it to 64-bit */
+	if (timo) {
+		struct __kernel_timespec ts_new;
+
+		ts_new.tv_sec = timo->tv_sec;
+		ts_new.tv_nsec = timo->tv_nsec;
+
+		return syscall(__NR_futex_waitv, waiters, nr_waiters, flags, &ts_new, clockid);
+	} else
+		return syscall(__NR_futex_waitv, waiters, nr_waiters, flags, NULL, clockid);
+}
+
 #endif /* _UAPI_LINUX_FUTEX_SYSCALL_H */
-- 
2.31.1