s390: Add infrastructure to patch lowcore accesses

The s390 architecture defines two special per-CPU data pages
called the "prefix area". In s390-linux terminology this is usually
called "lowcore". This memory area contains system configuration
data like old/new PSW's for system call/interrupt/machine check
handlers and lots of other data. It is normally mapped to logical
address 0. This area can only be accessed when in supervisor mode.

This means that kernel code can dereference NULL pointers, because
accesses to address 0 are allowed. Parts of lowcore can be write
protected, but read accesses and write accesses outside of the write
protected areas are not caught.

To remove this limitation for debugging and testing, remap lowcore to
another address and define a function get_lowcore() which simply
returns the address where lowcore is mapped at. This would normally
introduce a pointer dereference (=memory read). As lowcore is used
for several very often used variables, add code to patch this function
during runtime, so we avoid the memory reads.

For C code get_lowcore() has to be used, for assembly code it is
the GET_LC macro. When using this macro/function a reference is added
to alternative patching. All these locations will be patched to the
actual lowcore location when the kernel is booted or a module is loaded.

To make debugging/bisecting problems easier, this patch adds all the
infrastructure but the lowcore address is still hardwired to 0. This
way the code can be converted on a per function basis, and the
functionality is enabled in a patch after all the functions have
been converted.

Note that this requires at least z16 because the old lpsw instruction
only allowed a 12 bit displacement. z16 introduced lpswey which allows
20 bits (signed), so the lowcore can effectively be mapped from
address 0 - 0x7e000. To use 0x7e000 as address, a 6 byte lgfi
instruction would have to be used in the alternative. To save two
bytes, llilh can be used, but this only allows to set bits 16-31 of
the address. In order to use the llilh instruction, use 0x70000 as
alternative lowcore address. This is still large enough to catch
NULL pointer dereferences into large arrays.
Reviewed-by: Heiko Carstens <hca@linux.ibm.com>
Signed-off-by: Sven Schnelle <svens@linux.ibm.com>
Signed-off-by: Vasily Gorbik <gor@linux.ibm.com>

arch/s390/boot/boot.h

@@ -91,8 +91,10 @@ extern char _end[], _decompressor_end[];
 extern unsigned char _compressed_start[];
 extern unsigned char _compressed_end[];
 extern struct vmlinux_info _vmlinux_info;
+
 #define vmlinux _vmlinux_info
 
+#define __lowcore_pa(x)		((unsigned long)(x) % sizeof(struct lowcore))
 #define __abs_lowcore_pa(x)	(((unsigned long)(x) - __abs_lowcore) % sizeof(struct lowcore))
 #define __kernel_va(x)		((void *)((unsigned long)(x) - __kaslr_offset_phys + __kaslr_offset))
 #define __kernel_pa(x)		((unsigned long)(x) - __kaslr_offset + __kaslr_offset_phys)

arch/s390/boot/ipl_parm.c

@@ -3,6 +3,7 @@
 #include <linux/init.h>
 #include <linux/ctype.h>
 #include <linux/pgtable.h>
+#include <asm/abs_lowcore.h>
 #include <asm/page-states.h>
 #include <asm/ebcdic.h>
 #include <asm/sclp.h>

arch/s390/boot/startup.c

@@ -30,6 +30,7 @@ unsigned long __bootdata_preserved(vmemmap_size);
 unsigned long __bootdata_preserved(MODULES_VADDR);
 unsigned long __bootdata_preserved(MODULES_END);
 unsigned long __bootdata_preserved(max_mappable);
+int __bootdata_preserved(relocate_lowcore);
 
 u64 __bootdata_preserved(stfle_fac_list[16]);
 struct oldmem_data __bootdata_preserved(oldmem_data);

arch/s390/boot/vmem.c

@@ -26,6 +26,7 @@ atomic_long_t __bootdata_preserved(direct_pages_count[PG_DIRECT_MAP_MAX]);
 enum populate_mode {
 	POPULATE_NONE,
 	POPULATE_DIRECT,
+	POPULATE_LOWCORE,
 	POPULATE_ABS_LOWCORE,
 	POPULATE_IDENTITY,
 	POPULATE_KERNEL,
@@ -242,6 +243,8 @@ static unsigned long _pa(unsigned long addr, unsigned long size, enum populate_m
 		return -1;
 	case POPULATE_DIRECT:
 		return addr;
+	case POPULATE_LOWCORE:
+		return __lowcore_pa(addr);
 	case POPULATE_ABS_LOWCORE:
 		return __abs_lowcore_pa(addr);
 	case POPULATE_KERNEL:
@@ -418,6 +421,7 @@ static void pgtable_populate(unsigned long addr, unsigned long end, enum populat
 
 void setup_vmem(unsigned long kernel_start, unsigned long kernel_end, unsigned long asce_limit)
 {
+	unsigned long lowcore_address = 0;
 	unsigned long start, end;
 	unsigned long asce_type;
 	unsigned long asce_bits;
@@ -455,12 +459,17 @@ void setup_vmem(unsigned long kernel_start, unsigned long kernel_end, unsigned l
 	__arch_set_page_dat((void *)swapper_pg_dir, 1UL << CRST_ALLOC_ORDER);
 	__arch_set_page_dat((void *)invalid_pg_dir, 1UL << CRST_ALLOC_ORDER);
 
+	if (relocate_lowcore)
+		lowcore_address = LOWCORE_ALT_ADDRESS;
+
 	/*
 	 * To allow prefixing the lowcore must be mapped with 4KB pages.
 	 * To prevent creation of a large page at address 0 first map
 	 * the lowcore and create the identity mapping only afterwards.
 	 */
-	pgtable_populate(0, sizeof(struct lowcore), POPULATE_DIRECT);
+	pgtable_populate(lowcore_address,
+			 lowcore_address + sizeof(struct lowcore),
+			 POPULATE_LOWCORE);
 	for_each_physmem_usable_range(i, &start, &end) {
 		pgtable_populate((unsigned long)__identity_va(start),
 				 (unsigned long)__identity_va(end),

arch/s390/include/asm/abs_lowcore.h

@@ -2,6 +2,7 @@
 #ifndef _ASM_S390_ABS_LOWCORE_H
 #define _ASM_S390_ABS_LOWCORE_H
 
+#include <asm/sections.h>
 #include <asm/lowcore.h>
 
 #define ABS_LOWCORE_MAP_SIZE	(NR_CPUS * sizeof(struct lowcore))
@@ -24,4 +25,11 @@ static inline void put_abs_lowcore(struct lowcore *lc)
 	put_cpu();
 }
 
+extern int __bootdata_preserved(relocate_lowcore);
+
+static inline int have_relocated_lowcore(void)
+{
+	return relocate_lowcore;
+}
+
 #endif /* _ASM_S390_ABS_LOWCORE_H */

arch/s390/include/asm/alternative.h

@@ -33,6 +33,7 @@
 
 #define ALT_TYPE_FACILITY	0
 #define ALT_TYPE_SPEC		1
+#define ALT_TYPE_LOWCORE	2
 
 #define ALT_DATA_SHIFT		0
 #define ALT_TYPE_SHIFT		20
@@ -50,6 +51,9 @@
 					 ALT_TYPE_SPEC << ALT_TYPE_SHIFT	| \
 					 (facility) << ALT_DATA_SHIFT)
 
+#define ALT_LOWCORE			(ALT_CTX_EARLY << ALT_CTX_SHIFT		| \
+					 ALT_TYPE_LOWCORE << ALT_TYPE_SHIFT)
+
 #ifndef __ASSEMBLY__
 
 #include <linux/types.h>

arch/s390/include/asm/lowcore.h

@@ -14,10 +14,15 @@
 #include <asm/ctlreg.h>
 #include <asm/cpu.h>
 #include <asm/types.h>
+#include <asm/alternative.h>
 
 #define LC_ORDER 1
 #define LC_PAGES 2
 
+#define LOWCORE_ALT_ADDRESS	_AC(0x70000, UL)
+
+#ifndef __ASSEMBLY__
+
 struct pgm_tdb {
 	u64 data[32];
 };
@@ -214,7 +219,14 @@ struct lowcore {
 
 static __always_inline struct lowcore *get_lowcore(void)
 {
-	return NULL;
+	struct lowcore *lc;
+
+	if (__is_defined(__DECOMPRESSOR))
+		return NULL;
+	asm(ALTERNATIVE("llilh %[lc],0", "llilh %[lc],%[alt]", ALT_LOWCORE)
+	    : [lc] "=d" (lc)
+	    : [alt] "i" (LOWCORE_ALT_ADDRESS >> 16));
+	return lc;
 }
 
 extern struct lowcore *lowcore_ptr[];
@@ -224,4 +236,13 @@ static inline void set_prefix(__u32 address)
 	asm volatile("spx %0" : : "Q" (address) : "memory");
 }
 
+#else /* __ASSEMBLY__ */
+
+.macro GET_LC reg
+	ALTERNATIVE "llilh	\reg,0",					\
+		__stringify(llilh	\reg, LOWCORE_ALT_ADDRESS >> 16),	\
+		ALT_LOWCORE
+.endm
+
+#endif /* __ASSEMBLY__ */
 #endif /* _ASM_S390_LOWCORE_H */

arch/s390/kernel/abs_lowcore.c

@@ -4,6 +4,7 @@
 #include <asm/abs_lowcore.h>
 
 unsigned long __bootdata_preserved(__abs_lowcore);
+int __bootdata_preserved(relocate_lowcore);
 
 int abs_lowcore_map(int cpu, struct lowcore *lc, bool alloc)
 {

arch/s390/kernel/alternative.c

@@ -2,6 +2,7 @@
 
 #include <linux/uaccess.h>
 #include <asm/nospec-branch.h>
+#include <asm/abs_lowcore.h>
 #include <asm/alternative.h>
 #include <asm/facility.h>
 
@@ -25,6 +26,9 @@ void __apply_alternatives(struct alt_instr *start, struct alt_instr *end, unsign
 		case ALT_TYPE_SPEC:
 			replace = nobp_enabled();
 			break;
+		case ALT_TYPE_LOWCORE:
+			replace = have_relocated_lowcore();
+			break;
 		default:
 			replace = false;
 		}

arch/s390/kernel/early.c

@@ -48,6 +48,7 @@ decompressor_handled_param(dfltcc);
 decompressor_handled_param(facilities);
 decompressor_handled_param(nokaslr);
 decompressor_handled_param(cmma);
+decompressor_handled_param(relocate_lowcore);
 #if IS_ENABLED(CONFIG_KVM)
 decompressor_handled_param(prot_virt);
 #endif

arch/s390/kernel/setup.c

@@ -889,6 +889,9 @@ void __init setup_arch(char **cmdline_p)
 	else
 		pr_info("Linux is running as a guest in 64-bit mode\n");
 
+	if (have_relocated_lowcore())
+		pr_info("Lowcore relocated to 0x%px\n", get_lowcore());
+
 	log_component_list();
 
 	/* Have one command line that is parsed and saved in /proc/cmdline */