linux-user/loongarch64: Add vdso

Requires a relatively recent binutils version in order to avoid
spurious R_LARCH_NONE relocations.  The presence of these relocs
are diagnosed by our gen-vdso tool.

Tested-by: Song Gao <gaosong@loongson.cn>
Reviewed-by: Song Gao <gaosong@loongson.cn>
Reviewed-by: Philippe Mathieu-Daudé <philmd@linaro.org>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>

linux-user/elfload.c

@@ -1197,6 +1197,8 @@ static void elf_core_copy_regs(target_elf_gregset_t *regs, const CPUPPCState *en
 
 #define elf_check_arch(x) ((x) == EM_LOONGARCH)
 
+#define VDSO_HEADER "vdso.c.inc"
+
 static inline void init_thread(struct target_pt_regs *regs,
                                struct image_info *infop)
 {

linux-user/loongarch64/Makefile.vdso

@@ -0,0 +1,11 @@
+include $(BUILD_DIR)/tests/tcg/loongarch64-linux-user/config-target.mak
+
+SUBDIR = $(SRC_PATH)/linux-user/loongarch64
+VPATH += $(SUBDIR)
+
+all: $(SUBDIR)/vdso.so
+
+$(SUBDIR)/vdso.so: vdso.S vdso.ld vdso-asmoffset.h
+	$(CC) -o $@ -nostdlib -shared -fpic -Wl,-h,linux-vdso.so.1 \
+	  -Wl,--build-id=sha1 -Wl,--hash-style=both \
+	  -Wl,--no-warn-rwx-segments -Wl,-T,$(SUBDIR)/vdso.ld $<

linux-user/loongarch64/meson.build

@@ -0,0 +1,4 @@
+vdso_inc = gen_vdso.process('vdso.so',
+                            extra_args: ['-r', '__vdso_rt_sigreturn'])
+
+linux_user_ss.add(when: 'TARGET_LOONGARCH64', if_true: vdso_inc)

linux-user/loongarch64/signal.c

@@ -10,9 +10,9 @@
 #include "user-internals.h"
 #include "signal-common.h"
 #include "linux-user/trace.h"
-
 #include "target/loongarch/internals.h"
 #include "target/loongarch/vec.h"
+#include "vdso-asmoffset.h"
 
 /* FP context was used */
 #define SC_USED_FP              (1 << 0)
@@ -24,6 +24,11 @@ struct target_sigcontext {
     uint64_t sc_extcontext[0]   QEMU_ALIGNED(16);
 };
 
+QEMU_BUILD_BUG_ON(sizeof(struct target_sigcontext) != sizeof_sigcontext);
+QEMU_BUILD_BUG_ON(offsetof(struct target_sigcontext, sc_pc)
+                  != offsetof_sigcontext_pc);
+QEMU_BUILD_BUG_ON(offsetof(struct target_sigcontext, sc_regs)
+                  != offsetof_sigcontext_gr);
 
 #define FPU_CTX_MAGIC           0x46505501
 #define FPU_CTX_ALIGN           8
@@ -33,6 +38,9 @@ struct target_fpu_context {
     uint32_t fcsr;
 } QEMU_ALIGNED(FPU_CTX_ALIGN);
 
+QEMU_BUILD_BUG_ON(offsetof(struct target_fpu_context, regs)
+                  != offsetof_fpucontext_fr);
+
 #define CONTEXT_INFO_ALIGN      16
 struct target_sctx_info {
     uint32_t magic;
@@ -40,6 +48,8 @@ struct target_sctx_info {
     uint64_t padding;
 } QEMU_ALIGNED(CONTEXT_INFO_ALIGN);
 
+QEMU_BUILD_BUG_ON(sizeof(struct target_sctx_info) != sizeof_sctx_info);
+
 struct target_ucontext {
     abi_ulong tuc_flags;
     abi_ptr tuc_link;
@@ -54,6 +64,11 @@ struct target_rt_sigframe {
     struct target_ucontext       rs_uc;
 };
 
+QEMU_BUILD_BUG_ON(sizeof(struct target_rt_sigframe)
+                  != sizeof_rt_sigframe);
+QEMU_BUILD_BUG_ON(offsetof(struct target_rt_sigframe, rs_uc.tuc_mcontext)
+                  != offsetof_sigcontext);
+
 /*
  * These two structures are not present in guest memory, are private
  * to the signal implementation, but are largely copied from the

linux-user/loongarch64/vdso-asmoffset.h

@@ -0,0 +1,8 @@
+#define sizeof_rt_sigframe         0x240
+#define sizeof_sigcontext          0x110
+#define sizeof_sctx_info           0x10
+
+#define offsetof_sigcontext        0x130
+#define offsetof_sigcontext_pc     0
+#define offsetof_sigcontext_gr     8
+#define offsetof_fpucontext_fr     0

linux-user/loongarch64/vdso.S

@@ -0,0 +1,130 @@
+/*
+ * Loongarch64 linux replacement vdso.
+ *
+ * Copyright 2023 Linaro, Ltd.
+ *
+ * SPDX-License-Identifier: GPL-2.0-or-later
+ */
+
+#include <asm/unistd.h>
+#include <asm/errno.h>
+#include "vdso-asmoffset.h"
+
+
+	.text
+
+.macro endf name
+	.globl	\name
+	.type	\name, @function
+	.size	\name, . - \name
+.endm
+
+.macro vdso_syscall name, nr
+\name:
+	li.w	$a7, \nr
+	syscall	0
+	jr	$ra
+endf	\name
+.endm
+
+	.cfi_startproc
+
+vdso_syscall __vdso_gettimeofday, __NR_gettimeofday
+vdso_syscall __vdso_clock_gettime, __NR_clock_gettime
+vdso_syscall __vdso_clock_getres, __NR_clock_getres
+vdso_syscall __vdso_getcpu, __NR_getcpu
+
+	.cfi_endproc
+
+/*
+ * Start the unwind info at least one instruction before the signal
+ * trampoline, because the unwinder will assume we are returning
+ * after a call site.
+ */
+
+	.cfi_startproc simple
+	.cfi_signal_frame
+
+#define B_GR	offsetof_sigcontext_gr
+#define B_FR	sizeof_sigcontext + sizeof_sctx_info + offsetof_fpucontext_fr
+
+	.cfi_def_cfa	2, offsetof_sigcontext
+
+	/* Return address */
+	.cfi_return_column 64
+	.cfi_offset	64, offsetof_sigcontext_pc	/* pc */
+
+	/* Integer registers */
+	.cfi_offset	1, B_GR + 1 * 8
+	.cfi_offset	2, B_GR + 2 * 8
+	.cfi_offset	3, B_GR + 3 * 8
+	.cfi_offset	4, B_GR + 4 * 8
+	.cfi_offset	5, B_GR + 5 * 8
+	.cfi_offset	6, B_GR + 6 * 8
+	.cfi_offset	7, B_GR + 7 * 8
+	.cfi_offset	8, B_GR + 8 * 8
+	.cfi_offset	9, B_GR + 9 * 8
+	.cfi_offset	10, B_GR + 10 * 8
+	.cfi_offset	11, B_GR + 11 * 8
+	.cfi_offset	12, B_GR + 12 * 8
+	.cfi_offset	13, B_GR + 13 * 8
+	.cfi_offset	14, B_GR + 14 * 8
+	.cfi_offset	15, B_GR + 15 * 8
+	.cfi_offset	16, B_GR + 16 * 8
+	.cfi_offset	17, B_GR + 17 * 8
+	.cfi_offset	18, B_GR + 18 * 8
+	.cfi_offset	19, B_GR + 19 * 8
+	.cfi_offset	20, B_GR + 20 * 8
+	.cfi_offset	21, B_GR + 21 * 8
+	.cfi_offset	22, B_GR + 22 * 8
+	.cfi_offset	23, B_GR + 23 * 8
+	.cfi_offset	24, B_GR + 24 * 8
+	.cfi_offset	25, B_GR + 25 * 8
+	.cfi_offset	26, B_GR + 26 * 8
+	.cfi_offset	27, B_GR + 27 * 8
+	.cfi_offset	28, B_GR + 28 * 8
+	.cfi_offset	29, B_GR + 29 * 8
+	.cfi_offset	30, B_GR + 30 * 8
+	.cfi_offset	31, B_GR + 31 * 8
+
+	/* Floating point registers */
+	.cfi_offset	32, B_FR + 0
+	.cfi_offset	33, B_FR + 1 * 8
+	.cfi_offset	34, B_FR + 2 * 8
+	.cfi_offset	35, B_FR + 3 * 8
+	.cfi_offset	36, B_FR + 4 * 8
+	.cfi_offset	37, B_FR + 5 * 8
+	.cfi_offset	38, B_FR + 6 * 8
+	.cfi_offset	39, B_FR + 7 * 8
+	.cfi_offset	40, B_FR + 8 * 8
+	.cfi_offset	41, B_FR + 9 * 8
+	.cfi_offset	42, B_FR + 10 * 8
+	.cfi_offset	43, B_FR + 11 * 8
+	.cfi_offset	44, B_FR + 12 * 8
+	.cfi_offset	45, B_FR + 13 * 8
+	.cfi_offset	46, B_FR + 14 * 8
+	.cfi_offset	47, B_FR + 15 * 8
+	.cfi_offset	48, B_FR + 16 * 8
+	.cfi_offset	49, B_FR + 17 * 8
+	.cfi_offset	50, B_FR + 18 * 8
+	.cfi_offset	51, B_FR + 19 * 8
+	.cfi_offset	52, B_FR + 20 * 8
+	.cfi_offset	53, B_FR + 21 * 8
+	.cfi_offset	54, B_FR + 22 * 8
+	.cfi_offset	55, B_FR + 23 * 8
+	.cfi_offset	56, B_FR + 24 * 8
+	.cfi_offset	57, B_FR + 25 * 8
+	.cfi_offset	58, B_FR + 26 * 8
+	.cfi_offset	59, B_FR + 27 * 8
+	.cfi_offset	60, B_FR + 28 * 8
+	.cfi_offset	61, B_FR + 29 * 8
+	.cfi_offset	62, B_FR + 30 * 8
+	.cfi_offset	63, B_FR + 31 * 8
+
+	nop
+
+__vdso_rt_sigreturn:
+	li.w	$a7, __NR_rt_sigreturn
+	syscall	0
+	.cfi_endproc
+endf __vdso_rt_sigreturn

linux-user/loongarch64/vdso.ld

@@ -0,0 +1,73 @@
+/*
+ * Linker script for linux loongarch64 replacement vdso.
+ *
+ * Copyright 2023 Linaro, Ltd.
+ *
+ * SPDX-License-Identifier: GPL-2.0-or-later
+ */
+
+VERSION {
+        LINUX_5.10 {
+        global:
+                __vdso_getcpu;
+                __vdso_clock_getres;
+                __vdso_clock_gettime;
+                __vdso_gettimeofday;
+                __vdso_rt_sigreturn;
+
+        local: *;
+        };
+}
+
+
+PHDRS {
+        phdr            PT_PHDR         FLAGS(4) PHDRS;
+        load            PT_LOAD         FLAGS(7) FILEHDR PHDRS;
+        dynamic         PT_DYNAMIC      FLAGS(4);
+        eh_frame_hdr    PT_GNU_EH_FRAME;
+        note            PT_NOTE         FLAGS(4);
+}
+
+SECTIONS {
+        /*
+         * We can't prelink to any address without knowing something about
+         * the virtual memory space of the host, since that leaks over into
+         * the available memory space of the guest.
+         */
+        . = SIZEOF_HEADERS;
+
+        /*
+         * The following, including the FILEHDRS and PHDRS, are modified
+         * when we relocate the binary.  We want them to be initially
+         * writable for the relocation; we'll force them read-only after.
+         */
+        .note           : { *(.note*) }         :load :note
+        .dynamic        : { *(.dynamic) }       :load :dynamic
+        .dynsym         : { *(.dynsym) }        :load
+        /*
+         * There ought not be any real read-write data.
+         * But since we manipulated the segment layout,
+         * we have to put these sections somewhere.
+         */
+        .data           : {
+                *(.data*)
+                *(.sdata*)
+                *(.got.plt) *(.got)
+                *(.gnu.linkonce.d.*)
+                *(.bss*)
+                *(.dynbss*)
+                *(.gnu.linkonce.b.*)
+        }
+
+        .rodata         : { *(.rodata*) }
+        .hash           : { *(.hash) }
+        .gnu.hash       : { *(.gnu.hash) }
+        .dynstr         : { *(.dynstr) }
+        .gnu.version    : { *(.gnu.version) }
+        .gnu.version_d  : { *(.gnu.version_d) }
+        .gnu.version_r  : { *(.gnu.version_r) }
+        .eh_frame_hdr   : { *(.eh_frame_hdr) }  :load :eh_frame_hdr
+        .eh_frame       : { *(.eh_frame) }      :load
+
+        .text           : { *(.text*) }         :load   =0xd503201f
+}

linux-user/meson.build

@@ -40,6 +40,7 @@ subdir('alpha')
 subdir('arm')
 subdir('hppa')
 subdir('i386')
+subdir('loongarch64')
 subdir('m68k')
 subdir('microblaze')
 subdir('mips64')