s390/kexec_file: Add kexec_file_load system call

This patch adds the kexec_file_load system call to s390 as well as the arch
specific functions common code requires to work. Loaders for the different
file types will be added later.

Signed-off-by: Philipp Rudo <prudo@linux.vnet.ibm.com>
Signed-off-by: Martin Schwidefsky <schwidefsky@de.ibm.com>

arch/s390/Kconfig

@@ -51,6 +51,19 @@ config KEXEC
 	def_bool y
 	select KEXEC_CORE
 
+config KEXEC_FILE
+	bool "kexec file based system call"
+	select KEXEC_CORE
+	select BUILD_BIN2C
+	depends on CRYPTO
+	depends on CRYPTO_SHA256
+	depends on CRYPTO_SHA256_S390
+	---help---
+	  This is new version of kexec system call. This system call is
+	  file based and takes file descriptors as system call argument
+	  for kernel and initramfs as opposed to list of segments as
+	  accepted by previous system call.
+
 config ARCH_HAS_KEXEC_PURGATORY
 	def_bool y
 	depends on KEXEC_FILE

arch/s390/configs/default_defconfig

@@ -719,3 +719,4 @@ CONFIG_APPLDATA_BASE=y
 CONFIG_KVM=m
 CONFIG_KVM_S390_UCONTROL=y
 CONFIG_VHOST_NET=m
+CONFIG_KEXEC_FILE=y

arch/s390/kernel/Makefile

@@ -82,6 +82,8 @@ obj-$(CONFIG_FUNCTION_TRACER)	+= mcount.o ftrace.o
 obj-$(CONFIG_CRASH_DUMP)	+= crash_dump.o
 obj-$(CONFIG_UPROBES)		+= uprobes.o
 
+obj-$(CONFIG_KEXEC_FILE)	+= machine_kexec_file.o
+
 obj-$(CONFIG_PERF_EVENTS)	+= perf_event.o perf_cpum_cf.o perf_cpum_sf.o
 obj-$(CONFIG_PERF_EVENTS)	+= perf_cpum_cf_events.o perf_regs.o
 

arch/s390/kernel/compat_wrapper.c

@@ -182,3 +182,4 @@ COMPAT_SYSCALL_WRAP6(copy_file_range, int, fd_in, loff_t __user *, off_in, int,
 COMPAT_SYSCALL_WRAP2(s390_guarded_storage, int, command, struct gs_cb *, gs_cb);
 COMPAT_SYSCALL_WRAP5(statx, int, dfd, const char __user *, path, unsigned, flags, unsigned, mask, struct statx __user *, buffer);
 COMPAT_SYSCALL_WRAP4(s390_sthyi, unsigned long, code, void __user *, info, u64 __user *, rc, unsigned long, flags);
+COMPAT_SYSCALL_WRAP5(kexec_file_load, int, kernel_fd, int, initrd_fd, unsigned long, cmdline_len, const char __user *, cmdline_ptr, unsigned long, flags)

arch/s390/kernel/machine_kexec_file.c

@@ -0,0 +1,126 @@
+// SPDX-License-Identifier: GPL-2.0
+/*
+ * s390 code for kexec_file_load system call
+ *
+ * Copyright IBM Corp. 2018
+ *
+ * Author(s): Philipp Rudo <prudo@linux.vnet.ibm.com>
+ */
+
+#include <linux/elf.h>
+#include <linux/kexec.h>
+#include <asm/setup.h>
+
+const struct kexec_file_ops * const kexec_file_loaders[] = {
+	NULL,
+};
+
+/*
+ * The kernel is loaded to a fixed location. Turn off kexec_locate_mem_hole
+ * and provide kbuf->mem by hand.
+ */
+int arch_kexec_walk_mem(struct kexec_buf *kbuf,
+			int (*func)(struct resource *, void *))
+{
+	return 1;
+}
+
+int arch_kexec_apply_relocations_add(struct purgatory_info *pi,
+				     Elf_Shdr *section,
+				     const Elf_Shdr *relsec,
+				     const Elf_Shdr *symtab)
+{
+	Elf_Rela *relas;
+	int i;
+
+	relas = (void *)pi->ehdr + relsec->sh_offset;
+
+	for (i = 0; i < relsec->sh_size / sizeof(*relas); i++) {
+		const Elf_Sym *sym;	/* symbol to relocate */
+		unsigned long addr;	/* final location after relocation */
+		unsigned long val;	/* relocated symbol value */
+		void *loc;		/* tmp location to modify */
+
+		sym = (void *)pi->ehdr + symtab->sh_offset;
+		sym += ELF64_R_SYM(relas[i].r_info);
+
+		if (sym->st_shndx == SHN_UNDEF)
+			return -ENOEXEC;
+
+		if (sym->st_shndx == SHN_COMMON)
+			return -ENOEXEC;
+
+		if (sym->st_shndx >= pi->ehdr->e_shnum &&
+		    sym->st_shndx != SHN_ABS)
+			return -ENOEXEC;
+
+		loc = pi->purgatory_buf;
+		loc += section->sh_offset;
+		loc += relas[i].r_offset;
+
+		val = sym->st_value;
+		if (sym->st_shndx != SHN_ABS)
+			val += pi->sechdrs[sym->st_shndx].sh_addr;
+		val += relas[i].r_addend;
+
+		addr = section->sh_addr + relas[i].r_offset;
+
+		switch (ELF64_R_TYPE(relas[i].r_info)) {
+		case R_390_8:		/* Direct 8 bit.   */
+			*(u8 *)loc = val;
+			break;
+		case R_390_12:		/* Direct 12 bit.  */
+			*(u16 *)loc &= 0xf000;
+			*(u16 *)loc |= val & 0xfff;
+			break;
+		case R_390_16:		/* Direct 16 bit.  */
+			*(u16 *)loc = val;
+			break;
+		case R_390_20:		/* Direct 20 bit.  */
+			*(u32 *)loc &= 0xf00000ff;
+			*(u32 *)loc |= (val & 0xfff) << 16;	/* DL */
+			*(u32 *)loc |= (val & 0xff000) >> 4;	/* DH */
+			break;
+		case R_390_32:		/* Direct 32 bit.  */
+			*(u32 *)loc = val;
+			break;
+		case R_390_64:		/* Direct 64 bit.  */
+			*(u64 *)loc = val;
+			break;
+		case R_390_PC16:	/* PC relative 16 bit.	*/
+			*(u16 *)loc = (val - addr);
+			break;
+		case R_390_PC16DBL:	/* PC relative 16 bit shifted by 1.  */
+			*(u16 *)loc = (val - addr) >> 1;
+			break;
+		case R_390_PC32DBL:	/* PC relative 32 bit shifted by 1.  */
+			*(u32 *)loc = (val - addr) >> 1;
+			break;
+		case R_390_PC32:	/* PC relative 32 bit.	*/
+			*(u32 *)loc = (val - addr);
+			break;
+		case R_390_PC64:	/* PC relative 64 bit.	*/
+			*(u64 *)loc = (val - addr);
+			break;
+		default:
+			break;
+		}
+	}
+	return 0;
+}
+
+int arch_kexec_kernel_image_probe(struct kimage *image, void *buf,
+				  unsigned long buf_len)
+{
+	/* A kernel must be at least large enough to contain head.S. During
+	 * load memory in head.S will be accessed, e.g. to register the next
+	 * command line. If the next kernel were smaller the current kernel
+	 * will panic at load.
+	 *
+	 * 0x11000 = sizeof(head.S)
+	 */
+	if (buf_len < 0x11000)
+		return -ENOEXEC;
+
+	return kexec_image_probe_default(image, buf, buf_len);
+}

arch/s390/kernel/syscalls/syscall.tbl

@@ -388,3 +388,4 @@
 378  common	s390_guarded_storage	sys_s390_guarded_storage	compat_sys_s390_guarded_storage
 379  common	statx			sys_statx			compat_sys_statx
 380  common	s390_sthyi		sys_s390_sthyi			compat_sys_s390_sthyi
+381  common	kexec_file_load		sys_kexec_file_load		compat_sys_kexec_file_load