x86/stackprotector/32: Make the canary into a regular percpu variable

On 32-bit kernels, the stackprotector canary is quite nasty -- it is
stored at %gs:(20), which is nasty because 32-bit kernels use %fs for
percpu storage.  It's even nastier because it means that whether %gs
contains userspace state or kernel state while running kernel code
depends on whether stackprotector is enabled (this is
CONFIG_X86_32_LAZY_GS), and this setting radically changes the way
that segment selectors work.  Supporting both variants is a
maintenance and testing mess.

Merely rearranging so that percpu and the stack canary
share the same segment would be messy as the 32-bit percpu address
layout isn't currently compatible with putting a variable at a fixed
offset.

Fortunately, GCC 8.1 added options that allow the stack canary to be
accessed as %fs:__stack_chk_guard, effectively turning it into an ordinary
percpu variable.  This lets us get rid of all of the code to manage the
stack canary GDT descriptor and the CONFIG_X86_32_LAZY_GS mess.

(That name is special.  We could use any symbol we want for the
 %fs-relative mode, but for CONFIG_SMP=n, gcc refuses to let us use any
 name other than __stack_chk_guard.)

Forcibly disable stackprotector on older compilers that don't support
the new options and turn the stack canary into a percpu variable. The
"lazy GS" approach is now used for all 32-bit configurations.

Also makes load_gs_index() work on 32-bit kernels. On 64-bit kernels,
it loads the GS selector and updates the user GSBASE accordingly. (This
is unchanged.) On 32-bit kernels, it loads the GS selector and updates
GSBASE, which is now always the user base. This means that the overall
effect is the same on 32-bit and 64-bit, which avoids some ifdeffery.

 [ bp: Massage commit message. ]

Signed-off-by: Andy Lutomirski <luto@kernel.org>
Signed-off-by: Borislav Petkov <bp@suse.de>
Link: https://lkml.kernel.org/r/c0ff7dba14041c7e5d1cae5d4df052f03759bef3.1613243844.git.luto@kernel.org

1 files changed, 8 insertions, 22 deletions

diff --git a/arch/x86/include/asm/segment.h b/arch/x86/include/asm/segment.h
index 7fdd4facfce7..72044026eb3c 100644
--- a/arch/x86/include/asm/segment.h
+++ b/arch/x86/include/asm/segment.h
@@ -95,7 +95,7 @@
  *
  *  26 - ESPFIX small SS
  *  27 - per-cpu			[ offset to per-cpu data area ]
- *  28 - stack_canary-20		[ for stack protector ]		<=== cacheline #8
+ *  28 - unused
  *  29 - unused
  *  30 - unused
  *  31 - TSS for double fault handler
@@ -118,7 +118,6 @@
 
 #define GDT_ENTRY_ESPFIX_SS		26
 #define GDT_ENTRY_PERCPU		27
-#define GDT_ENTRY_STACK_CANARY		28
 
 #define GDT_ENTRY_DOUBLEFAULT_TSS	31
 
@@ -158,12 +157,6 @@
 # define __KERNEL_PERCPU		0
 #endif
 
-#ifdef CONFIG_STACKPROTECTOR
-# define __KERNEL_STACK_CANARY		(GDT_ENTRY_STACK_CANARY*8)
-#else
-# define __KERNEL_STACK_CANARY		0
-#endif
-
 #else /* 64-bit: */
 
 #include <asm/cache.h>
@@ -364,22 +357,15 @@ static inline void __loadsegment_fs(unsigned short value)
 	asm("mov %%" #seg ",%0":"=r" (value) : : "memory")
 
 /*
- * x86-32 user GS accessors:
+ * x86-32 user GS accessors.  This is ugly and could do with some cleaning up.
  */
 #ifdef CONFIG_X86_32
-# ifdef CONFIG_X86_32_LAZY_GS
-#  define get_user_gs(regs)		(u16)({ unsigned long v; savesegment(gs, v); v; })
-#  define set_user_gs(regs, v)		loadsegment(gs, (unsigned long)(v))
-#  define task_user_gs(tsk)		((tsk)->thread.gs)
-#  define lazy_save_gs(v)		savesegment(gs, (v))
-#  define lazy_load_gs(v)		loadsegment(gs, (v))
-# else	/* X86_32_LAZY_GS */
-#  define get_user_gs(regs)		(u16)((regs)->gs)
-#  define set_user_gs(regs, v)		do { (regs)->gs = (v); } while (0)
-#  define task_user_gs(tsk)		(task_pt_regs(tsk)->gs)
-#  define lazy_save_gs(v)		do { } while (0)
-#  define lazy_load_gs(v)		do { } while (0)
-# endif	/* X86_32_LAZY_GS */
+# define get_user_gs(regs)		(u16)({ unsigned long v; savesegment(gs, v); v; })
+# define set_user_gs(regs, v)		loadsegment(gs, (unsigned long)(v))
+# define task_user_gs(tsk)		((tsk)->thread.gs)
+# define lazy_save_gs(v)		savesegment(gs, (v))
+# define lazy_load_gs(v)		loadsegment(gs, (v))
+# define load_gs_index(v)		loadsegment(gs, (v))
 #endif	/* X86_32 */
 
 #endif /* !__ASSEMBLY__ */