Merge patch series "riscv: Introduce KASLR"

Alexandre Ghiti <alexghiti@rivosinc.com> says:

The following KASLR implementation allows to randomize the kernel mapping:

- virtually: we expect the bootloader to provide a seed in the device-tree
- physically: only implemented in the EFI stub, it relies on the firmware to
  provide a seed using EFI_RNG_PROTOCOL. arm64 has a similar implementation
  hence the patch 3 factorizes KASLR related functions for riscv to take
  advantage.

The new virtual kernel location is limited by the early page table that only
has one PUD and with the PMD alignment constraint, the kernel can only take
< 512 positions.

* b4-shazam-merge:
  riscv: libstub: Implement KASLR by using generic functions
  libstub: Fix compilation warning for rv32
  arm64: libstub: Move KASLR handling functions to kaslr.c
  riscv: Dump out kernel offset information on panic
  riscv: Introduce virtual kernel mapping KASLR

Link: https://lore.kernel.org/r/20230722123850.634544-1-alexghiti@rivosinc.com
Signed-off-by: Palmer Dabbelt <palmer@rivosinc.com>

10 files changed, 131 insertions, 2 deletions

diff --git a/arch/arm64/include/asm/efi.h b/arch/arm64/include/asm/efi.h
index f482b994c608..bcd5622aa096 100644
--- a/arch/arm64/include/asm/efi.h
+++ b/arch/arm64/include/asm/efi.h
@@ -156,4 +156,6 @@ static inline void efi_capsule_flush_cache_range(void *addr, int size)
 
 efi_status_t efi_handle_corrupted_x18(efi_status_t s, const char *f);
 
+void efi_icache_sync(unsigned long start, unsigned long end);
+
 #endif /* _ASM_EFI_H */
diff --git a/arch/riscv/Kconfig b/arch/riscv/Kconfig
index 468063780a93..712ae0269131 100644
--- a/arch/riscv/Kconfig
+++ b/arch/riscv/Kconfig
@@ -720,6 +720,25 @@ config RELOCATABLE
 
           If unsure, say N.
 
+config RANDOMIZE_BASE
+        bool "Randomize the address of the kernel image"
+        select RELOCATABLE
+        depends on MMU && 64BIT && !XIP_KERNEL
+        help
+          Randomizes the virtual address at which the kernel image is
+          loaded, as a security feature that deters exploit attempts
+          relying on knowledge of the location of kernel internals.
+
+          It is the bootloader's job to provide entropy, by passing a
+          random u64 value in /chosen/kaslr-seed at kernel entry.
+
+          When booting via the UEFI stub, it will invoke the firmware's
+          EFI_RNG_PROTOCOL implementation (if available) to supply entropy
+          to the kernel proper. In addition, it will randomise the physical
+          location of the kernel Image as well.
+
+          If unsure, say N.
+
 endmenu # "Kernel features"
 
 menu "Boot options"
diff --git a/arch/riscv/include/asm/efi.h b/arch/riscv/include/asm/efi.h
index 8a6a128ec57f..46a355913b27 100644
--- a/arch/riscv/include/asm/efi.h
+++ b/arch/riscv/include/asm/efi.h
@@ -45,4 +45,6 @@ void arch_efi_call_virt_teardown(void);
 
 unsigned long stext_offset(void);
 
+void efi_icache_sync(unsigned long start, unsigned long end);
+
 #endif /* _ASM_EFI_H */
diff --git a/arch/riscv/include/asm/page.h b/arch/riscv/include/asm/page.h
index b55ba20903ec..5488ecc337b6 100644
--- a/arch/riscv/include/asm/page.h
+++ b/arch/riscv/include/asm/page.h
@@ -106,6 +106,7 @@ typedef struct page *pgtable_t;
 struct kernel_mapping {
 	unsigned long page_offset;
 	unsigned long virt_addr;
+	unsigned long virt_offset;
 	uintptr_t phys_addr;
 	uintptr_t size;
 	/* Offset between linear mapping virtual address and kernel load address */
@@ -185,6 +186,8 @@ extern phys_addr_t __phys_addr_symbol(unsigned long x);
 
 #define sym_to_pfn(x)           __phys_to_pfn(__pa_symbol(x))
 
+unsigned long kaslr_offset(void);
+
 #endif /* __ASSEMBLY__ */
 
 #define virt_addr_valid(vaddr)	({						\
diff --git a/arch/riscv/kernel/image-vars.h b/arch/riscv/kernel/image-vars.h
index 15616155008c..ea1a10355ce9 100644
--- a/arch/riscv/kernel/image-vars.h
+++ b/arch/riscv/kernel/image-vars.h
@@ -27,6 +27,7 @@ __efistub__start		= _start;
 __efistub__start_kernel		= _start_kernel;
 __efistub__end			= _end;
 __efistub__edata		= _edata;
+__efistub___init_text_end	= __init_text_end;
 __efistub_screen_info		= screen_info;
 
 #endif
diff --git a/arch/riscv/kernel/pi/Makefile b/arch/riscv/kernel/pi/Makefile
index 7b593d44c712..07915dc9279e 100644
--- a/arch/riscv/kernel/pi/Makefile
+++ b/arch/riscv/kernel/pi/Makefile
@@ -35,5 +35,5 @@ $(obj)/string.o: $(srctree)/lib/string.c FORCE
 $(obj)/ctype.o: $(srctree)/lib/ctype.c FORCE
 	$(call if_changed_rule,cc_o_c)
 
-obj-y		:= cmdline_early.pi.o string.pi.o ctype.pi.o lib-fdt.pi.o lib-fdt_ro.pi.o
+obj-y		:= cmdline_early.pi.o fdt_early.pi.o string.pi.o ctype.pi.o lib-fdt.pi.o lib-fdt_ro.pi.o
 extra-y		:= $(patsubst %.pi.o,%.o,$(obj-y))
diff --git a/arch/riscv/kernel/pi/cmdline_early.c b/arch/riscv/kernel/pi/cmdline_early.c
index 05652d13c746..68e786c84c94 100644
--- a/arch/riscv/kernel/pi/cmdline_early.c
+++ b/arch/riscv/kernel/pi/cmdline_early.c
@@ -14,6 +14,7 @@ static char early_cmdline[COMMAND_LINE_SIZE];
  * LLVM complain because the function is actually unused in this file).
  */
 u64 set_satp_mode_from_cmdline(uintptr_t dtb_pa);
+bool set_nokaslr_from_cmdline(uintptr_t dtb_pa);
 
 static char *get_early_cmdline(uintptr_t dtb_pa)
 {
@@ -60,3 +61,15 @@ u64 set_satp_mode_from_cmdline(uintptr_t dtb_pa)
 
 	return match_noXlvl(cmdline);
 }
+
+static bool match_nokaslr(char *cmdline)
+{
+	return strstr(cmdline, "nokaslr");
+}
+
+bool set_nokaslr_from_cmdline(uintptr_t dtb_pa)
+{
+	char *cmdline = get_early_cmdline(dtb_pa);
+
+	return match_nokaslr(cmdline);
+}
diff --git a/arch/riscv/kernel/pi/fdt_early.c b/arch/riscv/kernel/pi/fdt_early.c
new file mode 100644
index 000000000000..899610e042ab
--- /dev/null
+++ b/arch/riscv/kernel/pi/fdt_early.c
@@ -0,0 +1,30 @@
+// SPDX-License-Identifier: GPL-2.0-only
+#include <linux/types.h>
+#include <linux/init.h>
+#include <linux/libfdt.h>
+
+/*
+ * Declare the functions that are exported (but prefixed) here so that LLVM
+ * does not complain it lacks the 'static' keyword (which, if added, makes
+ * LLVM complain because the function is actually unused in this file).
+ */
+u64 get_kaslr_seed(uintptr_t dtb_pa);
+
+u64 get_kaslr_seed(uintptr_t dtb_pa)
+{
+	int node, len;
+	fdt64_t *prop;
+	u64 ret;
+
+	node = fdt_path_offset((void *)dtb_pa, "/chosen");
+	if (node < 0)
+		return 0;
+
+	prop = fdt_getprop_w((void *)dtb_pa, node, "kaslr-seed", &len);
+	if (!prop || len != sizeof(u64))
+		return 0;
+
+	ret = fdt64_to_cpu(*prop);
+	*prop = 0;
+	return ret;
+}
diff --git a/arch/riscv/kernel/setup.c b/arch/riscv/kernel/setup.c
index 32c2e1eb71bd..e600aab116a4 100644
--- a/arch/riscv/kernel/setup.c
+++ b/arch/riscv/kernel/setup.c
@@ -21,6 +21,7 @@
 #include <linux/smp.h>
 #include <linux/efi.h>
 #include <linux/crash_dump.h>
+#include <linux/panic_notifier.h>
 
 #include <asm/acpi.h>
 #include <asm/alternative.h>
@@ -347,3 +348,27 @@ void free_initmem(void)
 
 	free_initmem_default(POISON_FREE_INITMEM);
 }
+
+static int dump_kernel_offset(struct notifier_block *self,
+			      unsigned long v, void *p)
+{
+	pr_emerg("Kernel Offset: 0x%lx from 0x%lx\n",
+		 kernel_map.virt_offset,
+		 KERNEL_LINK_ADDR);
+
+	return 0;
+}
+
+static struct notifier_block kernel_offset_notifier = {
+	.notifier_call = dump_kernel_offset
+};
+
+static int __init register_kernel_offset_dumper(void)
+{
+	if (IS_ENABLED(CONFIG_RANDOMIZE_BASE))
+		atomic_notifier_chain_register(&panic_notifier_list,
+					       &kernel_offset_notifier);
+
+	return 0;
+}
+device_initcall(register_kernel_offset_dumper);
diff --git a/arch/riscv/mm/init.c b/arch/riscv/mm/init.c
index 943c18d6ef4d..0798bd861dcb 100644
--- a/arch/riscv/mm/init.c
+++ b/arch/riscv/mm/init.c
@@ -1014,11 +1014,45 @@ static void __init pt_ops_set_late(void)
 #endif
 }
 
+#ifdef CONFIG_RANDOMIZE_BASE
+extern bool __init __pi_set_nokaslr_from_cmdline(uintptr_t dtb_pa);
+extern u64 __init __pi_get_kaslr_seed(uintptr_t dtb_pa);
+
+static int __init print_nokaslr(char *p)
+{
+	pr_info("Disabled KASLR");
+	return 0;
+}
+early_param("nokaslr", print_nokaslr);
+
+unsigned long kaslr_offset(void)
+{
+	return kernel_map.virt_offset;
+}
+#endif
+
 asmlinkage void __init setup_vm(uintptr_t dtb_pa)
 {
 	pmd_t __maybe_unused fix_bmap_spmd, fix_bmap_epmd;
 
-	kernel_map.virt_addr = KERNEL_LINK_ADDR;
+#ifdef CONFIG_RANDOMIZE_BASE
+	if (!__pi_set_nokaslr_from_cmdline(dtb_pa)) {
+		u64 kaslr_seed = __pi_get_kaslr_seed(dtb_pa);
+		u32 kernel_size = (uintptr_t)(&_end) - (uintptr_t)(&_start);
+		u32 nr_pos;
+
+		/*
+		 * Compute the number of positions available: we are limited
+		 * by the early page table that only has one PUD and we must
+		 * be aligned on PMD_SIZE.
+		 */
+		nr_pos = (PUD_SIZE - kernel_size) / PMD_SIZE;
+
+		kernel_map.virt_offset = (kaslr_seed % nr_pos) * PMD_SIZE;
+	}
+#endif
+
+	kernel_map.virt_addr = KERNEL_LINK_ADDR + kernel_map.virt_offset;
 	kernel_map.page_offset = _AC(CONFIG_PAGE_OFFSET, UL);
 
 #ifdef CONFIG_XIP_KERNEL