x86/syscalls: Revert "x86/syscalls: Make __X32_SYSCALL_BIT be unsigned long"
authorAndy Lutomirski <luto@kernel.org>
Sat, 9 May 2020 00:25:32 +0000 (17:25 -0700)
committerBorislav Petkov <bp@suse.de>
Tue, 26 May 2020 14:42:43 +0000 (16:42 +0200)
Revert

  45e29d119e99 ("x86/syscalls: Make __X32_SYSCALL_BIT be unsigned long")

and add a comment to discourage someone else from making the same
mistake again.

It turns out that some user code fails to compile if __X32_SYSCALL_BIT
is unsigned long. See, for example [1] below.

 [ bp: Massage and do the same thing in the respective tools/ header. ]

Fixes: 45e29d119e99 ("x86/syscalls: Make __X32_SYSCALL_BIT be unsigned long")
Reported-by: Thorsten Glaser <t.glaser@tarent.de>
Signed-off-by: Andy Lutomirski <luto@kernel.org>
Signed-off-by: Borislav Petkov <bp@suse.de>
Cc: stable@kernel.org
Link: [1] https://bugs.debian.org/cgi-bin/bugreport.cgi?bug=954294
Link: https://lkml.kernel.org/r/92e55442b744a5951fdc9cfee10badd0a5f7f828.1588983892.git.luto@kernel.org
arch/x86/include/uapi/asm/unistd.h
tools/arch/x86/include/uapi/asm/unistd.h

index 196fdd02b8b1b3d71ac994715f7dee8fa8b5fa52..be5e2e747f507657efc74f5ed2b68ed262103fda 100644 (file)
@@ -2,8 +2,15 @@
 #ifndef _UAPI_ASM_X86_UNISTD_H
 #define _UAPI_ASM_X86_UNISTD_H
 
-/* x32 syscall flag bit */
-#define __X32_SYSCALL_BIT      0x40000000UL
+/*
+ * x32 syscall flag bit.  Some user programs expect syscall NR macros
+ * and __X32_SYSCALL_BIT to have type int, even though syscall numbers
+ * are, for practical purposes, unsigned long.
+ *
+ * Fortunately, expressions like (nr & ~__X32_SYSCALL_BIT) do the right
+ * thing regardless.
+ */
+#define __X32_SYSCALL_BIT      0x40000000
 
 #ifndef __KERNEL__
 # ifdef __i386__
index 196fdd02b8b1b3d71ac994715f7dee8fa8b5fa52..30d7d04d72d6bad6dfb419f7606e8d5c8fd9b622 100644 (file)
@@ -3,7 +3,7 @@
 #define _UAPI_ASM_X86_UNISTD_H
 
 /* x32 syscall flag bit */
-#define __X32_SYSCALL_BIT      0x40000000UL
+#define __X32_SYSCALL_BIT      0x40000000
 
 #ifndef __KERNEL__
 # ifdef __i386__