rseq: Make rseq work with protection keys

index 2e6c04d8a45b4..614099766d5f2 100644
@@ -2,6 +2,8 @@
 #ifndef _ASM_X86_PKEYS_H
 #define _ASM_X86_PKEYS_H
 
+#include "pkru.h"
+
 /*
  * If more than 16 keys are ever supported, a thorough audit
  * will be necessary to ensure that the types that store key

@@ -123,4 +125,32 @@ static inline int vma_pkey(struct vm_area_struct *vma)
 	return (vma->vm_flags & vma_pkey_mask) >> VM_PKEY_SHIFT;
 }
 
+typedef u32 pkey_reg_t;
+
+static inline pkey_reg_t write_permissive_pkey_val(void)
+{
+	return write_pkru(0);
+}
+
+static inline pkey_reg_t enable_zero_pkey_val(void)
+{
+	u32 pkru;
+
+	if (!cpu_feature_enabled(X86_FEATURE_OSPKE))
+		return 0;
+	/*
+	 * WRPKRU is relatively expensive compared to RDPKRU,
+	 * avoid it if possible.
+	 */
+	pkru = rdpkru();
+	if ((pkru & (PKRU_AD_BIT|PKRU_WD_BIT)) != 0)
+		wrpkru(pkru & ~(PKRU_AD_BIT|PKRU_WD_BIT));
+	return pkru;
+}
+
+static inline void write_pkey_val(pkey_reg_t val)
+{
+	write_pkru(val);
+}
+
 #endif /*_ASM_X86_PKEYS_H */

index 74f0a2d34ffdd..b9bf9b7f2753b 100644
@@ -39,16 +39,20 @@ static inline u32 read_pkru(void)
 	return 0;
 }
 
-static inline void write_pkru(u32 pkru)
+static inline u32 write_pkru(u32 pkru)
 {
+	u32 old_pkru;
+
 	if (!cpu_feature_enabled(X86_FEATURE_OSPKE))
-		return;
+		return 0;
 	/*
 	 * WRPKRU is relatively expensive compared to RDPKRU.
 	 * Avoid WRPKRU when it would not change the value.
 	 */
-	if (pkru != rdpkru())
+	old_pkru = rdpkru();
+	if (pkru != old_pkru)
 		wrpkru(pkru);
+	return old_pkru;
 }
 
 static inline void pkru_write_default(void)

index 86be8bf27b41b..262d60f6a15f8 100644
@@ -48,4 +48,35 @@ static inline bool arch_pkeys_enabled(void)
 
 #endif /* ! CONFIG_ARCH_HAS_PKEYS */
 
+#ifndef CONFIG_ARCH_HAS_PERMISSIVE_PKEY
+
+/*
+ * Common name for value of the register that controls access to PKEYs
+ * (called differently on different arches: PKRU, POR, AMR).
+ */
+typedef char pkey_reg_t;
+
+/*
+ * Sets PKEY access register to the most permissive value that allows
+ * accesses to all PKEYs. Returns the current value of PKEY register.
+ * Code should generally arrange switching back to the old value
+ * using write_pkey_val(old_value).
+ */
+static inline pkey_reg_t write_permissive_pkey_val(void)
+{
+	return 0;
+}
+
+/*
+ * Sets PKEY access register to a value that allows access to the 0 (default)
+ * PKEY. Returns the current value of PKEY register.
+ */
+static inline pkey_reg_t enable_zero_pkey_val(void)
+{
+	return 0;
+}
+
+static inline void write_pkey_val(pkey_reg_t val) {}
+#endif /* ! CONFIG_ARCH_HAS_PERMISSIVE_PKEY */
+
 #endif /* _LINUX_PKEYS_H */

index 5f441039b5725..27a66a0697dd2 100644
@@ -28,6 +28,7 @@
 #include <linux/entry-common.h>
 #include <linux/syscalls.h>
 #include <linux/rseq.h>
+#include <linux/pkeys.h>
 
 #include <asm/processor.h>
 #include <asm/ucontext.h>

@@ -72,10 +73,7 @@ static inline int is_x32_frame(struct ksignal *ksig)
  */
 static inline u32 sig_prepare_pkru(void)
 {
-	u32 orig_pkru = read_pkru();
-
-	write_pkru(0);
-	return orig_pkru;
+	return write_permissive_pkey_val();
 }
 
 /*

index 2cb16091ec0ae..9d9c976d3b78c 100644
@@ -10,6 +10,7 @@
 
 #include <linux/sched.h>
 #include <linux/uaccess.h>
+#include <linux/pkeys.h>
 #include <linux/syscalls.h>
 #include <linux/rseq.h>
 #include <linux/types.h>

@@ -402,11 +403,19 @@ static int rseq_ip_fixup(struct pt_regs *regs)
 void __rseq_handle_notify_resume(struct ksignal *ksig, struct pt_regs *regs)
 {
 	struct task_struct *t = current;
+	pkey_reg_t saved_pkey;
 	int ret, sig;
 
 	if (unlikely(t->flags & PF_EXITING))
 		return;
 
+	/*
+	 * Enable access to the default (0) pkey in case the thread has
+	 * currently disabled access to it and struct rseq/rseq_cs has
+	 * 0 pkey assigned (the only supported value for now).
+	 */
+	saved_pkey = enable_zero_pkey_val();
+
 	/*
 	 * regs is NULL if and only if the caller is in a syscall path.  Skip
 	 * fixup and leave rseq_cs as is so that rseq_sycall() will detect and

@@ -419,9 +428,11 @@ void __rseq_handle_notify_resume(struct ksignal *ksig, struct pt_regs *regs)
 	}
 	if (unlikely(rseq_update_cpu_node_id(t)))
 		goto error;
+	write_pkey_val(saved_pkey);
 	return;
 
 error:
+	write_pkey_val(saved_pkey);
 	sig = ksig ? ksig->sig : 0;
 	force_sigsegv(sig);
 }

98 lines not on this page · raw message

index ba424ce80a719..65da4a727c550 100644
@@ -8,6 +8,7 @@
 #ifndef RSEQ_H
 #define RSEQ_H
 
+#include <assert.h>
 #include <stdint.h>
 #include <stdbool.h>
 #include <pthread.h>