tools/nolibc: shrink arch support
Subsystems (3)
v3 cover tools/nolibc: shrink arch support
Hi, Willy, Thomas
Here is the revision of the v2 arch support shrink patchset [1], it mainly applies suggestions from you.
It is based on the 20230710-nolibc-ser2-tom-syscall-configv4-report branch of nolibc repo.
Tested for all of the supported archs:
arch/board | result
------------|------------
arm/versatilepb | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.
arm/vexpress-a9 | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.
arm/virt | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.
aarch64/virt | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.
i386/pc | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.
x86_64/pc | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.
mipsel/malta | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.
loongarch64/virt | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.
riscv64/virt | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.
riscv32/virt | 151 test(s): 122 passed, 7 skipped, 22 failed => status: failure.
s390x/s390-ccw-virtio | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.Changes from v2 --> v3:
* tools/nolibc: remove the old sys_stat support
Revert the reorg operation, basically the same as v1
* tools/nolibc: add new crt.h with _start_c
Revert the reorg operation
Add crt.h in Makefile
Add _nolibc_main alias for main to silence the compile warning about
* tools/nolibc: arm: shrink _start with _start_c tools/nolibc: aarch64: shrink _start with _start_c tools/nolibc: i386: shrink _start with _start_c tools/nolibc: x86_64: shrink _start with _start_c tools/nolibc: mips: shrink _start with _start_c tools/nolibc: loongarch: shrink _start with _start_c tools/nolibc: riscv: shrink _start with _start_c tools/nolibc: s390: shrink _start with _start_c
Revert the reorg operation, but still use post-whitespaces instead of post-tab. Include "crt.h" in arch-<ARCH>.h
* tools/nolibc: arch-*.h: add missing space after ','
Fix up the errors reported by scripts/checkpatch.pl
Best regards,
Zhangjin
---
[1]: https://lore.kernel.org/lkml/[email protected]/
Zhangjin Wu (11):
tools/nolibc: remove the old sys_stat support tools/nolibc: add new crt.h with _start_c tools/nolibc: arm: shrink _start with _start_c tools/nolibc: aarch64: shrink _start with _start_c tools/nolibc: i386: shrink _start with _start_c tools/nolibc: x86_64: shrink _start with _start_c tools/nolibc: mips: shrink _start with _start_c tools/nolibc: loongarch: shrink _start with _start_c tools/nolibc: riscv: shrink _start with _start_c tools/nolibc: s390: shrink _start with _start_c tools/nolibc: arch-*.h: add missing space after ','
tools/include/nolibc/Makefile | 1 + tools/include/nolibc/arch-aarch64.h | 56 ++---------------- tools/include/nolibc/arch-arm.h | 82 +++------------------------ tools/include/nolibc/arch-i386.h | 61 +++----------------- tools/include/nolibc/arch-loongarch.h | 45 ++------------- tools/include/nolibc/arch-mips.h | 76 ++++--------------------- tools/include/nolibc/arch-riscv.h | 68 +++------------------- tools/include/nolibc/arch-s390.h | 63 ++------------------ tools/include/nolibc/arch-x86_64.h | 57 +++---------------- tools/include/nolibc/crt.h | 59 +++++++++++++++++++ tools/include/nolibc/sys.h | 63 ++++---------------- tools/include/nolibc/types.h | 4 +- 12 files changed, 133 insertions(+), 502 deletions(-) create mode 100644 tools/include/nolibc/crt.h
Signature · 1 line
2.25.1
PATCH v3 1/11 tools/nolibc: remove the old sys_stat support
__NR_statx has been added from v4.10:
commit a528d35e8bfc ("statx: Add a system call to make enhanced file info available")
It has been supported by all of the platforms since at least from v4.20 and glibc 2.28.
Let's remove the old arch related and dependent sys_stat support completely.
This is friendly to the future new architecture porting.
--- tools/include/nolibc/arch-aarch64.h | 28 ------------- tools/include/nolibc/arch-arm.h | 37 ----------------- tools/include/nolibc/arch-i386.h | 26 ------------ tools/include/nolibc/arch-mips.h | 28 ------------- tools/include/nolibc/arch-riscv.h | 23 ----------- tools/include/nolibc/arch-s390.h | 25 ------------ tools/include/nolibc/arch-x86_64.h | 27 ------------- tools/include/nolibc/sys.h | 63 +++++------------------------ tools/include/nolibc/types.h | 4 +- 9 files changed, 13 insertions(+), 248 deletions(-)
tools/include/nolibc/arch-aarch64.h +0 -28
@@ -9,34 +9,6 @@#include "compiler.h" -/* The struct returned by the newfstatat() syscall. Differs slightly from the -* x86_64's stat one by field ordering, so be careful. -*/ -struct sys_stat_struct { -unsigned long st_dev; -unsigned long st_ino; -unsigned int st_mode; -unsigned int st_nlink; -unsigned int st_uid; -unsigned int st_gid; - -unsigned long st_rdev; -unsigned long __pad1; -long st_size; -int st_blksize; -int __pad2; - -long st_blocks; -long st_atime; -unsigned long st_atime_nsec; -long st_mtime; - -unsigned long st_mtime_nsec; -long st_ctime; -unsigned long st_ctime_nsec; -unsigned int __unused[2 ]; -}; -/* Syscalls for AARCH64 : * - registers are 64-bit * - stack is 16-byte aligned
tools/include/nolibc/arch-arm.h +0 -37
@@ -9,43 +9,6 @@#include "compiler.h" -/* The struct returned by the stat() syscall, 32-bit only, the syscall returns -* exactly 56 bytes (stops before the unused array). In big endian, the format -* differs as devices are returned as short only. -*/ -struct sys_stat_struct { -#if defined(__ARMEB__) -unsigned short st_dev; -unsigned short __pad1; -#else -unsigned long st_dev; -#endif -unsigned long st_ino; -unsigned short st_mode; -unsigned short st_nlink; -unsigned short st_uid; -unsigned short st_gid; - -#if defined(__ARMEB__) -unsigned short st_rdev; -unsigned short __pad2; -#else -unsigned long st_rdev; -#endif -unsigned long st_size; -unsigned long st_blksize; -unsigned long st_blocks; - -unsigned long st_atime; -unsigned long st_atime_nsec; -unsigned long st_mtime; -unsigned long st_mtime_nsec; - -unsigned long st_ctime; -unsigned long st_ctime_nsec; -unsigned long __unused[2 ]; -}; -/* Syscalls for ARM in ARM or Thumb modes : * - registers are 32-bit * - stack is 8-byte aligned
tools/include/nolibc/arch-i386.h +0 -26
@@ -9,32 +9,6 @@#include "compiler.h" -/* The struct returned by the stat() syscall, 32-bit only, the syscall returns -* exactly 56 bytes (stops before the unused array). -*/ -struct sys_stat_struct { -unsigned long st_dev; -unsigned long st_ino; -unsigned short st_mode; -unsigned short st_nlink; -unsigned short st_uid; -unsigned short st_gid; - -unsigned long st_rdev; -unsigned long st_size; -unsigned long st_blksize; -unsigned long st_blocks; - -unsigned long st_atime; -unsigned long st_atime_nsec; -unsigned long st_mtime; -unsigned long st_mtime_nsec; - -unsigned long st_ctime; -unsigned long st_ctime_nsec; -unsigned long __unused[2 ]; -}; -/* Syscalls for i386 : * - mostly similar to x86_64 * - registers are 32-bit
tools/include/nolibc/arch-mips.h +0 -28
@@ -9,34 +9,6 @@#include "compiler.h" -/* The struct returned by the stat() syscall. 88 bytes are returned by the -* syscall. -*/ -struct sys_stat_struct { -unsigned int st_dev; -long st_pad1[3 ]; -unsigned long st_ino; -unsigned int st_mode; -unsigned int st_nlink; -unsigned int st_uid; -unsigned int st_gid; -unsigned int st_rdev; -long st_pad2[2 ]; -long st_size; -long st_pad3; - -long st_atime; -long st_atime_nsec; -long st_mtime; -long st_mtime_nsec; - -long st_ctime; -long st_ctime_nsec; -long st_blksize; -long st_blocks; -long st_pad4[14 ]; -}; -/* Syscalls for MIPS ABI O32 : * - WARNING! there's always a delayed slot! * - WARNING again, the syntax is different, registers take a '$' and numbers
tools/include/nolibc/arch-riscv.h +0 -23
@@ -9,29 +9,6 @@#include "compiler.h" -struct sys_stat_struct { -unsigned long st_dev;/* Device. */ -unsigned long st_ino;/* File serial number. */ -unsigned int st_mode;/* File mode. */ -unsigned int st_nlink;/* Link count. */ -unsigned int st_uid;/* User ID of the file's owner. */ -unsigned int st_gid;/* Group ID of the file's group. */ -unsigned long st_rdev;/* Device number, if device. */ -unsigned long __pad1; -long st_size;/* Size of file, in bytes. */ -int st_blksize;/* Optimal block size for I/O. */ -int __pad2; -long st_blocks;/* Number 512-byte blocks allocated. */ -long st_atime;/* Time of last access. */ -unsigned long st_atime_nsec; -long st_mtime;/* Time of last modification. */ -unsigned long st_mtime_nsec; -long st_ctime;/* Time of last status change. */ -unsigned long st_ctime_nsec; -unsigned int __unused4; -unsigned int __unused5; -}; -#if __riscv_xlen ==64 #define PTRLOG "3"#define SZREG "8"
tools/include/nolibc/arch-s390.h +0 -25
@@ -10,31 +10,6 @@#include "compiler.h" -/* The struct returned by the stat() syscall, equivalent to stat64(). The -* syscall returns 116 bytes and stops in the middle of __unused. -*/ - -struct sys_stat_struct { -unsigned long st_dev; -unsigned long st_ino; -unsigned long st_nlink; -unsigned int st_mode; -unsigned int st_uid; -unsigned int st_gid; -unsigned int __pad1; -unsigned long st_rdev; -unsigned long st_size; -unsigned long st_atime; -unsigned long st_atime_nsec; -unsigned long st_mtime; -unsigned long st_mtime_nsec; -unsigned long st_ctime; -unsigned long st_ctime_nsec; -unsigned long st_blksize; -long st_blocks; -unsigned long __unused[3 ]; -}; -/* Syscalls for s390: * - registers are 64-bit * - syscall number is passed in r1
tools/include/nolibc/arch-x86_64.h +0 -27
@@ -9,33 +9,6 @@#include "compiler.h" -/* The struct returned by the stat() syscall, equivalent to stat64(). The -* syscall returns 116 bytes and stops in the middle of __unused. -*/ -struct sys_stat_struct { -unsigned long st_dev; -unsigned long st_ino; -unsigned long st_nlink; -unsigned int st_mode; -unsigned int st_uid; - -unsigned int st_gid; -unsigned int __pad0; -unsigned long st_rdev; -long st_size; -long st_blksize; - -long st_blocks; -unsigned long st_atime; -unsigned long st_atime_nsec; -unsigned long st_mtime; - -unsigned long st_mtime_nsec; -unsigned long st_ctime; -unsigned long st_ctime_nsec; -long __unused[3 ]; -}; -/* Syscalls for x86_64 : * - registers are 64-bit * - syscall number is passed in rax
tools/include/nolibc/sys.h +11 -52
97 lines not on this page · raw message
tools/include/nolibc/types.h +2 -2
@@ -15,8 +15,8 @@/* Only the generic macros and types may be defined here. The arch-specific -* ones such as the O_RDONLY and related macros used by fcntl() and open(), or -* the layout of sys_stat_struct must not be defined here. +* ones such as the O_RDONLY and related macros used by fcntl() and open() +* must not be defined here. */ /* stat flags (WARNING, octal here). We need to check for an existing
Signature · 1 line
2.25.1
PATCH v3 2/11 tools/nolibc: add new crt.h with _start_c
In reply to Zhangjin WuAs the environ and _auxv support added for nolibc, the assembly _start function becomes more and more complex and therefore makes the porting of nolibc to new architectures harder and harder.
To simplify portability, this C version of _start_c() is added to do most of the assembly start operations in C, which reduces the complexity a lot and will eventually simplify the porting of nolibc to the new architectures.
The new _start_c() only requires a stack pointer argument, it will find argv, envp and _auxv for us, and then call main(), finally, it exit() with main's return status. With this new _start_c(), the future new architectures only require to add very few assembly instructions.
As suggested by Thomas, users may use a different signature of main (e.g. void main(void)), a _nolibc_main alias is added for main to silence the warning about potential conflicting types.
--- tools/include/nolibc/Makefile | 1 + tools/include/nolibc/crt.h | 59 +++++++++++++++++++++++++++++++++++ 2 files changed, 60 insertions(+) create mode 100644 tools/include/nolibc/crt.h
tools/include/nolibc/Makefile +1 -0
@@ -27,6 +27,7 @@ nolibc_arch := $(patsubst arm64,aarch64,$(ARCH)) arch_file := arch-$(nolibc_arch).h all_files := \ compiler.h \ + crt.h \ ctype.h \ errno.h \ nolibc.h \
tools/include/nolibc/crt.h new file +59 -0
59 lines not on this page · raw message
Signature · 1 line
2.25.1
PATCH v3 3/11 tools/nolibc: arm: shrink _start with _start_c
In reply to Zhangjin Wumove most of the _start operations to _start_c().
--- tools/include/nolibc/arch-arm.h | 43 +++++---------------------------- 1 file changed, 6 insertions(+), 37 deletions(-)
tools/include/nolibc/arch-arm.h +6 -37
@@ -8,6 +8,7 @@#define _NOLIBC_ARCH_ARM_H#include "compiler.h" +#include "crt.h" /* Syscalls for ARM in ARM or Thumb modes : * - registers are 32-bit
@@ -183,49 +184,17 @@ _arg1; \ }) - -char **environ __attribute__((weak)); -const unsigned long *_auxv __attribute__((weak)); -/* startup code */ void __attribute__((weak,noreturn,optimize("omit-frame-pointer"))) __no_stack_protector _start(void) { __asm__volatile ( #ifdef _NOLIBC_STACKPROTECTOR - "bl __stack_chk_init\n" /* initialize stack protector */ + "bl __stack_chk_init\n" /* initialize stack protector */ #endif -"pop {%r0}\n" /* argc was in the stack */ -"mov %r1, %sp\n" /* argv = sp */ - -"add %r2, %r0, $1\n" /* envp = (argc + 1) ... */ -"lsl %r2, %r2, $2\n" /* * 4 ... */ -"add %r2, %r2, %r1\n" /* + argv */ -"ldr %r3, 1f\n" /* r3 = &environ (see below) */ -"str %r2, [r3]\n" /* store envp into environ */ - -"mov r4, r2\n" /* search for auxv (follows NULL after last env) */ -"0:\n" -"mov r5, r4\n" /* r5 = r4 */ -"add r4, r4, #4\n" /* r4 += 4 */ -"ldr r5,[r5]\n" /* r5 = *r5 = *(r4-4) */ -"cmp r5, #0\n" /* and stop at NULL after last env */ -"bne 0b\n" -"ldr %r3, 2f\n" /* r3 = &_auxv (low bits) */ -"str r4, [r3]\n" /* store r4 into _auxv */ - -"mov %r3, $8\n" /* AAPCS : sp must be 8-byte aligned in the */ -"neg %r3, %r3\n" /* callee, and bl doesn't push (lr=pc) */ -"and %r3, %r3, %r1\n" /* so we do sp = r1(=sp) & r3(=-8); */ -"mov %sp, %r3\n" - -"bl main\n" /* main() returns the status code, we'll exit with it. */ -"movs r7, $1\n" /* NR_exit == 1 */ -"svc $0x00\n" -".align 2\n" /* below are the pointers to a few variables */ -"1:\n" -".word environ\n" -"2:\n" -".word _auxv\n" +"mov %r0, sp\n" /* save stack pointer to %r0, as arg1 of _start_c */ +"and ip, %r0, #-8\n" /* sp must be 8-byte aligned in the callee */ +"mov sp, ip\n" +"bl _start_c\n" /* transfer to c runtime */ );__builtin_unreachable (); }
Signature · 1 line
2.25.1
PATCH v3 4/11 tools/nolibc: aarch64: shrink _start with _start_c
In reply to Zhangjin Wumove most of the _start operations to _start_c().
--- tools/include/nolibc/arch-aarch64.h | 26 +++++--------------------- 1 file changed, 5 insertions(+), 21 deletions(-)
tools/include/nolibc/arch-aarch64.h +5 -21
@@ -8,6 +8,7 @@#define _NOLIBC_ARCH_AARCH64_H#include "compiler.h" +#include "crt.h" /* Syscalls for AARCH64 : * - registers are 64-bit
@@ -143,33 +144,16 @@ _arg1; \ }) -char **environ __attribute__((weak)); -const unsigned long *_auxv __attribute__((weak)); -/* startup code */ void __attribute__((weak,noreturn,optimize("omit-frame-pointer"))) __no_stack_protector _start(void) { __asm__volatile ( #ifdef _NOLIBC_STACKPROTECTOR - "bl __stack_chk_init\n" /* initialize stack protector */ + "bl __stack_chk_init\n" /* initialize stack protector */ #endif -"ldr x0, [sp]\n" /* argc (x0) was in the stack */ -"add x1, sp, 8\n" /* argv (x1) = sp */ -"lsl x2, x0, 3\n" /* envp (x2) = 8*argc ... */ -"add x2, x2, 8\n" /* + 8 (skip null) */ -"add x2, x2, x1\n" /* + argv */ -"adrp x3, environ\n" /* x3 = &environ (high bits) */ -"str x2, [x3, #:lo12:environ]\n" /* store envp into environ */ -"mov x4, x2\n" /* search for auxv (follows NULL after last env) */ -"0:\n" -"ldr x5, [x4], 8\n" /* x5 = *x4; x4 += 8 */ -"cbnz x5, 0b\n" /* and stop at NULL after last env */ -"adrp x3, _auxv\n" /* x3 = &_auxv (high bits) */ -"str x4, [x3, #:lo12:_auxv]\n" /* store x4 into _auxv */ -"and sp, x1, -16\n" /* sp must be 16-byte aligned in the callee */ -"bl main\n" /* main() returns the status code, we'll exit with it. */ -"mov x8, 93\n" /* NR_exit == 93 */ -"svc #0\n" +"mov x0, sp\n" /* save stack pointer to x0, as arg1 of _start_c */ +"and sp, x0, -16\n" /* sp must be 16-byte aligned in the callee */ +"bl _start_c\n" /* transfer to c runtime */ );__builtin_unreachable (); }
Signature · 1 line
2.25.1
PATCH v3 5/11 tools/nolibc: i386: shrink _start with _start_c
In reply to Zhangjin Wumove most of the _start operations to _start_c().
--- tools/include/nolibc/arch-i386.h | 33 ++++++++------------------------ 1 file changed, 8 insertions(+), 25 deletions(-)
tools/include/nolibc/arch-i386.h +8 -25
@@ -8,6 +8,7 @@#define _NOLIBC_ARCH_I386_H#include "compiler.h" +#include "crt.h" /* Syscalls for i386 : * - mostly similar to x86_64
@@ -154,9 +155,6 @@ _eax; \ }) -char **environ __attribute__((weak)); -const unsigned long *_auxv __attribute__((weak)); -/* startup code */ /* * i386 System V ABI mandates:
@@ -168,29 +166,14 @@ void __attribute__((weak,noreturn,optimize("omit-frame-pointer"))) __no_stack_pr { __asm__volatile ( #ifdef _NOLIBC_STACKPROTECTOR - "call __stack_chk_init\n" /* initialize stack protector */ + "call __stack_chk_init\n" /* initialize stack protector */ #endif -"pop %eax\n" /* argc (first arg, %eax) */ -"mov %esp, %ebx\n" /* argv[] (second arg, %ebx) */ -"lea 4(%ebx,%eax,4),%ecx\n" /* then a NULL then envp (third arg, %ecx) */ -"mov %ecx, environ\n" /* save environ */ -"xor %ebp, %ebp\n" /* zero the stack frame */ -"mov %ecx, %edx\n" /* search for auxv (follows NULL after last env) */ -"0:\n" -"add $4, %edx\n" /* search for auxv using edx, it follows the */ -"cmp -4(%edx), %ebp\n" /* ... NULL after last env (ebp is zero here) */ -"jnz 0b\n" -"mov %edx, _auxv\n" /* save it into _auxv */ -"and $-16, %esp\n" /* x86 ABI : esp must be 16-byte aligned before */ -"sub $4, %esp\n" /* the call instruction (args are aligned) */ -"push %ecx\n" /* push all registers on the stack so that we */ -"push %ebx\n" /* support both regparm and plain stack modes */ -"push %eax\n" -"call main\n" /* main() returns the status code in %eax */ -"mov %eax, %ebx\n" /* retrieve exit code (32-bit int) */ -"movl $1, %eax\n" /* NR_exit == 1 */ -"int $0x80\n" /* exit now */ -"hlt\n" /* ensure it does not */ +"xor %ebp, %ebp\n" /* zero the stack frame */ +"mov %esp, %eax\n" /* save stack pointer to %eax, as arg1 of _start_c */ +"and $-16, %esp\n" /* last pushed argument must be 16-byte aligned */ +"push %eax\n" /* push arg1 on stack to support plain stack modes too */ +"call _start_c\n" /* transfer to c runtime */ +"hlt\n" /* ensure it does not return */ );__builtin_unreachable (); }
Signature · 1 line
2.25.1
PATCH v3 7/11 tools/nolibc: mips: shrink _start with _start_c
In reply to Zhangjin Wumove most of the _start operations to _start_c().
Also clean up the instructions in delayed slots.
--- tools/include/nolibc/arch-mips.h | 46 +++++++------------------------- 1 file changed, 10 insertions(+), 36 deletions(-)
tools/include/nolibc/arch-mips.h +10 -36
@@ -8,6 +8,7 @@#define _NOLIBC_ARCH_MIPS_H#include "compiler.h" +#include "crt.h" /* Syscalls for MIPS ABI O32 : * - WARNING! there's always a delayed slot!
@@ -173,50 +174,23 @@ _arg4 ? -_num : _num; \ }) -char **environ __attribute__((weak)); -const unsigned long *_auxv __attribute__((weak)); -/* startup code, note that it's called __start on MIPS */ void __attribute__((weak,noreturn,optimize("omit-frame-pointer"))) __no_stack_protector __start(void) { __asm__volatile ( -/*".set nomips16\n"*/ ".set push\n" - ".set noreorder\n" + ".set noreorder\n" ".option pic0\n" #ifdef _NOLIBC_STACKPROTECTOR - "jal __stack_chk_init\n" /* initialize stack protector */ - "nop\n" /* delayed slot */ + "jal __stack_chk_init\n" /* initialize stack protector */ + " nop\n" /* delayed slot */ #endif -/*".ent __start\n"*/ -/*"__start:\n"*/ -"lw $a0,($sp)\n" /* argc was in the stack */ -"addiu $a1, $sp, 4\n" /* argv = sp + 4 */ -"sll $a2, $a0, 2\n" /* a2 = argc * 4 */ -"add $a2, $a2, $a1\n" /* envp = argv + 4*argc ... */ -"addiu $a2, $a2, 4\n" /* ... + 4 */ -"lui $a3, %hi(environ)\n" /* load environ into a3 (hi) */ -"addiu $a3, %lo(environ)\n" /* load environ into a3 (lo) */ -"sw $a2,($a3)\n" /* store envp(a2) into environ */ - -"move $t0, $a2\n" /* iterate t0 over envp, look for NULL */ -"0:" /* do { */ -"lw $a3, ($t0)\n" /* a3=*(t0); */ -"bne $a3, $0, 0b\n" /* } while (a3); */ -"addiu $t0, $t0, 4\n" /* delayed slot: t0+=4; */ -"lui $a3, %hi(_auxv)\n" /* load _auxv into a3 (hi) */ -"addiu $a3, %lo(_auxv)\n" /* load _auxv into a3 (lo) */ -"sw $t0, ($a3)\n" /* store t0 into _auxv */ - -"li $t0, -8\n" -"and $sp, $sp, $t0\n" /* sp must be 8-byte aligned */ -"addiu $sp,$sp,-16\n" /* the callee expects to save a0..a3 there! */ -"jal main\n" /* main() returns the status code, we'll exit with it. */ -"nop\n" /* delayed slot */ -"move $a0, $v0\n" /* retrieve 32-bit exit code from v0 */ -"li $v0, 4001\n" /* NR_exit == 4001 */ -"syscall\n" -/*".end __start\n"*/ +"move $a0, $sp\n" /* save stack pointer to $a0, as arg1 of _start_c */ +"li $t0, -8\n" +"and $sp, $sp, $t0\n" /* $sp must be 8-byte aligned */ +"addiu $sp, $sp, -16\n" /* the callee expects to save a0..a3 there */ +"jal _start_c\n" /* transfer to c runtime */ +" nop\n" /* delayed slot */ ".set pop\n" );__builtin_unreachable ();
Signature · 1 line
2.25.1
PATCH v3 8/11 tools/nolibc: loongarch: shrink _start with _start_c
In reply to Zhangjin Wumove most of the _start operations to _start_c().
--- tools/include/nolibc/arch-loongarch.h | 43 ++++----------------------- 1 file changed, 5 insertions(+), 38 deletions(-)
tools/include/nolibc/arch-loongarch.h +5 -38
@@ -8,6 +8,7 @@#define _NOLIBC_ARCH_LOONGARCH_H#include "compiler.h" +#include "crt.h" /* Syscalls for LoongArch : * - stack is 16-byte aligned
@@ -143,26 +144,9 @@ _arg1; \ }) -char **environ __attribute__((weak)); -const unsigned long *_auxv __attribute__((weak)); -#if __loongarch_grlen ==32 -#define LONGLOG "2" -#define SZREG "4" -#define REG_L "ld.w" -#define LONG_S "st.w" -#define LONG_ADD "add.w" -#define LONG_ADDI "addi.w" -#define LONG_SLL "slli.w"#define LONG_BSTRINS "bstrins.w"#else /* __loongarch_grlen == 64 */ -#define LONGLOG "3" -#define SZREG "8" -#define REG_L "ld.d" -#define LONG_S "st.d" -#define LONG_ADD "add.d" -#define LONG_ADDI "addi.d" -#define LONG_SLL "slli.d"#define LONG_BSTRINS "bstrins.d"#endif
@@ -171,28 +155,11 @@ void __attribute__((weak,noreturn,optimize("omit-frame-pointer"))) __no_stack_pr { __asm__volatile ( #ifdef _NOLIBC_STACKPROTECTOR - "bl __stack_chk_init\n" /* initialize stack protector */ + "bl __stack_chk_init\n" /* initialize stack protector */ #endif -REG_L " $a0, $sp, 0\n" /* argc (a0) was in the stack */ -LONG_ADDI " $a1, $sp, " SZREG "\n" /* argv (a1) = sp + SZREG */ -LONG_SLL " $a2, $a0, " LONGLOG "\n" /* envp (a2) = SZREG*argc ... */ -LONG_ADDI " $a2, $a2, " SZREG "\n" /* + SZREG (skip null) */ -LONG_ADD " $a2, $a2, $a1\n" /* + argv */ - -"move $a3, $a2\n" /* iterate a3 over envp to find auxv (after NULL) */ -"0:\n" /* do { */ -REG_L " $a4, $a3, 0\n" /* a4 = *a3; */ -LONG_ADDI " $a3, $a3, " SZREG "\n" /* a3 += sizeof(void*); */ -"bne $a4, $zero, 0b\n" /* } while (a4); */ -"la.pcrel $a4, _auxv\n" /* a4 = &_auxv */ -LONG_S " $a3, $a4, 0\n" /* store a3 into _auxv */ - -"la.pcrel $a3, environ\n" /* a3 = &environ */ -LONG_S " $a2, $a3, 0\n" /* store envp(a2) into environ */ -LONG_BSTRINS " $sp, $zero, 3, 0\n" /* sp must be 16-byte aligned */ -"bl main\n" /* main() returns the status code, we'll exit with it. */ -"li.w $a7, 93\n" /* NR_exit == 93 */ -"syscall 0\n" +"move $a0, $sp\n" /* save stack pointer to $a0, as arg1 of _start_c */ +LONG_BSTRINS " $sp, $zero, 3, 0\n" /* $sp must be 16-byte aligned */ +"bl _start_c\n" /* transfer to c runtime */ );__builtin_unreachable (); }
Signature · 1 line
2.25.1
PATCH v3 9/11 tools/nolibc: riscv: shrink _start with _start_c
In reply to Zhangjin Wumove most of the _start operations to _start_c().
--- tools/include/nolibc/arch-riscv.h | 43 +++++-------------------------- 1 file changed, 6 insertions(+), 37 deletions(-)
tools/include/nolibc/arch-riscv.h +6 -37
@@ -8,18 +8,7 @@#define _NOLIBC_ARCH_RISCV_H#include "compiler.h" - -#if __riscv_xlen ==64 -#define PTRLOG "3" -#define SZREG "8" -#define REG_L "ld" -#define REG_S "sd" -#elif __riscv_xlen ==32 -#define PTRLOG "2" -#define SZREG "4" -#define REG_L "lw" -#define REG_S "sw" -#endif +#include "crt.h" /* Syscalls for RISCV : * - stack is 16-byte aligned
@@ -153,40 +142,20 @@ _arg1; \ }) -char **environ __attribute__((weak)); -const unsigned long *_auxv __attribute__((weak)); -/* startup code */ void __attribute__((weak,noreturn,optimize("omit-frame-pointer"))) __no_stack_protector _start(void) { __asm__volatile ( ".option push\n" ".option norelax\n" - "lla gp, __global_pointer$\n" + "lla gp, __global_pointer$\n" ".option pop\n" #ifdef _NOLIBC_STACKPROTECTOR - "call __stack_chk_init\n" /* initialize stack protector */ + "call __stack_chk_init\n" /* initialize stack protector */ #endif -REG_L " a0, 0(sp)\n" /* argc (a0) was in the stack */ -"add a1, sp, " SZREG "\n" /* argv (a1) = sp */ -"slli a2, a0, " PTRLOG "\n" /* envp (a2) = SZREG*argc ... */ -"add a2, a2, " SZREG "\n" /* + SZREG (skip null) */ -"add a2,a2,a1\n" /* + argv */ - -"add a3, a2, zero\n" /* iterate a3 over envp to find auxv (after NULL) */ -"0:\n" /* do { */ -REG_L " a4, 0(a3)\n" /* a4 = *a3; */ -"add a3, a3, " SZREG "\n" /* a3 += sizeof(void*); */ -"bne a4, zero, 0b\n" /* } while (a4); */ -"lui a4, %hi(_auxv)\n" /* a4 = &_auxv (high bits) */ -REG_S " a3, %lo(_auxv)(a4)\n" /* store a3 into _auxv */ - -"lui a3, %hi(environ)\n" /* a3 = &environ (high bits) */ -REG_S " a2,%lo(environ)(a3)\n" /* store envp(a2) into environ */ -"andi sp,a1,-16\n" /* sp must be 16-byte aligned */ -"call main\n" /* main() returns the status code, we'll exit with it. */ -"li a7, 93\n" /* NR_exit == 93 */ -"ecall\n" +"mv a0, sp\n" /* save stack pointer to a0, as arg1 of _start_c */ +"andi sp, a0, -16\n" /* sp must be 16-byte aligned */ +"call _start_c\n" /* transfer to c runtime */ );__builtin_unreachable (); }
Signature · 1 line
2.25.1
PATCH v3 10/11 tools/nolibc: s390: shrink _start with _start_c
In reply to Zhangjin Wumove most of the _start operations to _start_c().
--- tools/include/nolibc/arch-s390.h | 36 +++++--------------------------- 1 file changed, 5 insertions(+), 31 deletions(-)
tools/include/nolibc/arch-s390.h +5 -31
@@ -9,6 +9,7 @@#include <asm/unistd.h> #include "compiler.h" +#include "crt.h" /* Syscalls for s390: * - registers are 64-bit
@@ -137,41 +138,14 @@ _arg1; \ }) -char **environ __attribute__((weak)); -const unsigned long *_auxv __attribute__((weak)); -/* startup code */ void __attribute__((weak,noreturn,optimize("omit-frame-pointer"))) __no_stack_protector _start(void) { __asm__volatile ( -"lg %r2,0(%r15)\n" /* argument count */ -"la %r3,8(%r15)\n" /* argument pointers */ - -"xgr %r0,%r0\n" /* r0 will be our NULL value */ -/* search for envp */ -"lgr %r4,%r3\n" /* start at argv */ -"0:\n" -"clg %r0,0(%r4)\n" /* entry zero? */ -"la %r4,8(%r4)\n" /* advance pointer */ -"jnz 0b\n" /* no -> test next pointer */ -/* yes -> r4 now contains start of envp */ -"larl %r1,environ\n" -"stg %r4,0(%r1)\n" - -/* search for auxv */ -"lgr %r5,%r4\n" /* start at envp */ -"1:\n" -"clg %r0,0(%r5)\n" /* entry zero? */ -"la %r5,8(%r5)\n" /* advance pointer */ -"jnz 1b\n" /* no -> test next pointer */ -"larl %r1,_auxv\n" /* yes -> store value in _auxv */ -"stg %r5,0(%r1)\n" - -"aghi %r15,-160\n" /* allocate new stackframe */ -"xc 0(8,%r15),0(%r15)\n" /* clear backchain */ -"brasl %r14,main\n" /* ret value of main is arg to exit */ -"lghi %r1,1\n" /* __NR_exit */ -"svc 0\n" +"lgr %r2, %r15\n" /* save stack pointer to %r2, as arg1 of _start_c */ +"aghi %r15, -160\n" /* allocate new stackframe */ +"xc 0(8,%r15), 0(%r15)\n" /* clear backchain */ +"brasl %r14, _start_c\n" /* transfer to c runtime */ );__builtin_unreachable (); }
Signature · 1 line
2.25.1
PATCH v3 11/11 tools/nolibc: arch-*.h: add missing space after ','
In reply to Zhangjin WuFix up such errors reported by scripts/checkpatch.pl:
ERROR: space required after that ',' (ctx:VxV)
#148: FILE: tools/include/nolibc/arch-aarch64.h:148:
+void __attribute__((weak,noreturn,optimize("omit-frame-pointer"))) __no_stack_protector _start(void)
^
ERROR: space required after that ',' (ctx:VxV)
#148: FILE: tools/include/nolibc/arch-aarch64.h:148:
+void __attribute__((weak,noreturn,optimize("omit-frame-pointer"))) __no_stack_protector _start(void)
^
--- tools/include/nolibc/arch-aarch64.h | 2 +- tools/include/nolibc/arch-arm.h | 2 +- tools/include/nolibc/arch-i386.h | 2 +- tools/include/nolibc/arch-loongarch.h | 2 +- tools/include/nolibc/arch-mips.h | 2 +- tools/include/nolibc/arch-riscv.h | 2 +- tools/include/nolibc/arch-s390.h | 2 +- tools/include/nolibc/arch-x86_64.h | 2 +- 8 files changed, 8 insertions(+), 8 deletions(-)
tools/include/nolibc/arch-aarch64.h +1 -1
@@ -145,7 +145,7 @@ })/* startup code */ -void __attribute__((weak,noreturn,optimize("omit-frame-pointer"))) __no_stack_protector _start(void) +void __attribute__((weak, noreturn, optimize("omit-frame-pointer"))) __no_stack_protector _start(void) { __asm__ volatile (#ifdef _NOLIBC_STACKPROTECTOR
tools/include/nolibc/arch-arm.h +1 -1
@@ -185,7 +185,7 @@ })/* startup code */ -void __attribute__((weak,noreturn,optimize("omit-frame-pointer"))) __no_stack_protector _start(void) +void __attribute__((weak, noreturn, optimize("omit-frame-pointer"))) __no_stack_protector _start(void) { __asm__ volatile (#ifdef _NOLIBC_STACKPROTECTOR
tools/include/nolibc/arch-i386.h +1 -1
@@ -162,7 +162,7 @@* 2) The deepest stack frame should be set to zero * */ -void __attribute__((weak,noreturn,optimize("omit-frame-pointer"))) __no_stack_protector _start(void) +void __attribute__((weak, noreturn, optimize("omit-frame-pointer"))) __no_stack_protector _start(void) { __asm__ volatile (#ifdef _NOLIBC_STACKPROTECTOR
tools/include/nolibc/arch-loongarch.h +1 -1
@@ -151,7 +151,7 @@#endif /* startup code */ -void __attribute__((weak,noreturn,optimize("omit-frame-pointer"))) __no_stack_protector _start(void) +void __attribute__((weak, noreturn, optimize("omit-frame-pointer"))) __no_stack_protector _start(void) { __asm__ volatile (#ifdef _NOLIBC_STACKPROTECTOR
tools/include/nolibc/arch-mips.h +1 -1
@@ -175,7 +175,7 @@ })/* startup code, note that it's called __start on MIPS */ -void __attribute__((weak,noreturn,optimize("omit-frame-pointer"))) __no_stack_protector __start(void) +void __attribute__((weak, noreturn, optimize("omit-frame-pointer"))) __no_stack_protector __start(void) { __asm__volatile ( ".set push\n"
tools/include/nolibc/arch-riscv.h +1 -1
@@ -143,7 +143,7 @@ })/* startup code */ -void __attribute__((weak,noreturn,optimize("omit-frame-pointer"))) __no_stack_protector _start(void) +void __attribute__((weak, noreturn, optimize("omit-frame-pointer"))) __no_stack_protector _start(void) { __asm__volatile ( ".option push\n"
tools/include/nolibc/arch-s390.h +1 -1
@@ -139,7 +139,7 @@ })/* startup code */ -void __attribute__((weak,noreturn,optimize("omit-frame-pointer"))) __no_stack_protector _start(void) +void __attribute__((weak, noreturn, optimize("omit-frame-pointer"))) __no_stack_protector _start(void) { __asm__volatile ( "lgr %r2, %r15\n" /* save stack pointer to %r2, as arg1 of _start_c */
tools/include/nolibc/arch-x86_64.h +1 -1
@@ -161,7 +161,7 @@* 2) The deepest stack frame should be zero (the %rbp). * */ -void __attribute__((weak,noreturn,optimize("omit-frame-pointer"))) __no_stack_protector _start(void) +void __attribute__((weak, noreturn, optimize("omit-frame-pointer"))) __no_stack_protector _start(void) { __asm__ volatile (#ifdef _NOLIBC_STACKPROTECTOR
Signature · 1 line
2.25.1
PATCH v3 6/11 tools/nolibc: x86_64: shrink _start with _start_c
In reply to Zhangjin Wumove most of the _start operations to _start_c().
--- tools/include/nolibc/arch-x86_64.h | 28 +++++++--------------------- 1 file changed, 7 insertions(+), 21 deletions(-)
tools/include/nolibc/arch-x86_64.h +7 -21
@@ -8,6 +8,7 @@#define _NOLIBC_ARCH_X86_64_H#include "compiler.h" +#include "crt.h" /* Syscalls for x86_64 : * - registers are 64-bit
@@ -153,9 +154,6 @@ _ret; \ }) -char **environ __attribute__((weak)); -const unsigned long *_auxv __attribute__((weak)); -/* startup code */ /* * x86-64 System V ABI mandates:
@@ -167,25 +165,13 @@ void __attribute__((weak,noreturn,optimize("omit-frame-pointer"))) __no_stack_pr { __asm__volatile ( #ifdef _NOLIBC_STACKPROTECTOR - "call __stack_chk_init\n" /* initialize stack protector */ + "call __stack_chk_init\n" /* initialize stack protector */ #endif -"pop %rdi\n" /* argc (first arg, %rdi) */ -"mov %rsp, %rsi\n" /* argv[] (second arg, %rsi) */ -"lea 8(%rsi,%rdi,8),%rdx\n" /* then a NULL then envp (third arg, %rdx) */ -"mov %rdx, environ\n" /* save environ */ -"xor %ebp, %ebp\n" /* zero the stack frame */ -"mov %rdx, %rax\n" /* search for auxv (follows NULL after last env) */ -"0:\n" -"add $8, %rax\n" /* search for auxv using rax, it follows the */ -"cmp -8(%rax), %rbp\n" /* ... NULL after last env (rbp is zero here) */ -"jnz 0b\n" -"mov %rax, _auxv\n" /* save it into _auxv */ -"and $-16, %rsp\n" /* x86 ABI : esp must be 16-byte aligned before call */ -"call main\n" /* main() returns the status code, we'll exit with it. */ -"mov %eax, %edi\n" /* retrieve exit code (32 bit) */ -"mov $60, %eax\n" /* NR_exit == 60 */ -"syscall\n" /* really exit */ -"hlt\n" /* ensure it does not return */ +"xor %ebp, %ebp\n" /* zero the stack frame */ +"mov %rsp, %rdi\n" /* save stack pointer to %rdi, as arg1 of _start_c */ +"and $-16, %rsp\n" /* %rsp must be 16-byte aligned before call */ +"call _start_c\n" /* transfer to c runtime */ +"hlt\n" /* ensure it does not return */ );__builtin_unreachable (); }
Signature · 1 line
2.25.1
Re: [PATCH v3 02/11] tools/nolibc: add new crt.h with _start_c
In reply to Zhangjin WuQuoted 48 lines · Zhangjin Wu
As the environ and
_auxvsupport added for nolibc, the assembly_startfunction becomes more and more complex and therefore makes the porting of nolibc to new architectures harder and harder.To simplify portability, this C version of
_start_c()is added to do most of the assembly start operations in C, which reduces the complexity a lot and will eventually simplify the porting of nolibc to the new architectures.The new
_start_c()only requires a stack pointer argument, it will find argv, envp and_auxvfor us, and then callmain(), finally, itexit()with main's return status. With this new_start_c(), the future new architectures only require to add very few assembly instructions.As suggested by Thomas, users may use a different signature of main (e.g. void main(void)), a
_nolibc_mainalias is added for main to silence the warning about potential conflicting types.Suggested-by Thomas Weißschuh <[email protected]>Signed-off-by Zhangjin Wu <[email protected]>--- tools/include/nolibc/Makefile | 1 + tools/include/nolibc/crt.h | 59 +++++++++++++++++++++++++++++++++++ 2 files changed, 60 insertions(+) create mode 100644 tools/include/nolibc/crt.htools/include/nolibc/Makefile +1
-0@@ -27,6 +27,7 @@ nolibc_arch := $(patsubst arm64,aarch64,$(ARCH)) arch_file := arch-$(nolibc_arch).h all_files := \ compiler.h \ + crt.h \ ctype.h \ errno.h \ nolibc.h \tools/include/nolibc/crt.h new file +3
-0@@ -0,0 +1,59 @@ +/* SPDX-License-Identifier: LGPL-2.1 OR MIT */ +/* +* C Run Time support for NOLIBC
tools/include/nolibc/crt.h line 13
Quoted 10 lines
tools/include/nolibc/crt.h new file +10
-0+* Copyright (C) 2023 Zhangjin Wu <[email protected]> +*/ + +#ifndef _NOLIBC_CRT_H +#define _NOLIBC_CRT_H + +char **environ __attribute__((weak)); +const unsigned long *_auxv __attribute__((weak)); + +typedef int (_nolibc_main_fn )(int ,char **,char **);
What's the advantage of the typedef over using the pointer type inline?
tools/include/nolibc/crt.h line 22
Quoted 9 lines
tools/include/nolibc/crt.h new file +9
-0+static void exit (int ); + +void _start_c (long *sp) +{ +int argc, i; +char **argv; +char **envp; +/* silence potential warning: conflicting types for 'main' */ +_nolibc_main_fn _nolibc_main __asm__ ("main" );
What about the stackprotector initialization?
It would really fit great into this series.
Quoted 11 lines
tools/include/nolibc/crt.h new file +11
-0+ +/* +* sp : argc <-- argument count, required by main() +* argv: argv[0] <-- argument vector, required by main() +* argv[1] +* ... +* argv[argc-1] +* null +* envp: envp[0] <-- environment variables, required by main() and getenv() +* envp[1] +* ...
tools/include/nolibc/crt.h line 43
Quoted 10 lines
tools/include/nolibc/crt.h new file +10
-0+* null +* _auxv: auxv[0] <-- auxiliary vector, required by getauxval() +* auxv[1] +* ... +* null +*/ + +/* assign argc and argv */ + argc = sp[0 ]; + argv = (void *)(sp +1 );
Bit of a weird mismatch between array syntax and pointer arithmetic.
tools/include/nolibc/crt.h line 47
Quoted 4 lines
tools/include/nolibc/crt.h new file +4
-0+ +/* find envp */ + envp = argv + argc +1 ; + environ = envp;
Is envp really needed? Could just be assigned directly to environ.
tools/include/nolibc/crt.h line 53
Quoted 6 lines
tools/include/nolibc/crt.h new file +6
-0+ +/* find auxv */ + i =0 ; +while (envp[i]) + i++; + _auxv = (void *)(envp + i +1 );
Could be simplified a bit:
_auxv = (void *) envp;
while (_auxv)
_auxv++;
Quoted 9 lines
tools/include/nolibc/crt.h new file +6
-0+ +/* go to application */ + exit(_nolibc_main(argc, argv, envp)); +} + +#endif /* _NOLIBC_CRT_H */ Signature · 1 line
2.25.1
Re: [PATCH v3 02/11] tools/nolibc: add new crt.h with _start_c
Hi, Thomas
Quoted 53 lines · Zhangjin Wu
Quoted 52 lines
As the environ and
_auxvsupport added for nolibc, the assembly_startfunction becomes more and more complex and therefore makes the porting of nolibc to new architectures harder and harder.To simplify portability, this C version of
_start_c()is added to do most of the assembly start operations in C, which reduces the complexity a lot and will eventually simplify the porting of nolibc to the new architectures.The new
_start_c()only requires a stack pointer argument, it will find argv, envp and_auxvfor us, and then callmain(), finally, itexit()with main's return status. With this new_start_c(), the future new architectures only require to add very few assembly instructions.As suggested by Thomas, users may use a different signature of main (e.g. void main(void)), a
_nolibc_mainalias is added for main to silence the warning about potential conflicting types.Suggested-by Thomas Weißschuh <[email protected]>Signed-off-by Zhangjin Wu <[email protected]>--- tools/include/nolibc/Makefile | 1 + tools/include/nolibc/crt.h | 59 +++++++++++++++++++++++++++++++++++ 2 files changed, 60 insertions(+) create mode 100644 tools/include/nolibc/crt.htools/include/nolibc/Makefile +1
-0@@ -27,6 +27,7 @@ nolibc_arch := $(patsubst arm64,aarch64,$(ARCH)) arch_file := arch-$(nolibc_arch).h all_files := \ compiler.h \ + crt.h \ ctype.h \ errno.h \ nolibc.h \tools/include/nolibc/crt.h new file +7
-0@@ -0,0 +1,59 @@ +/* SPDX-License-Identifier: LGPL-2.1 OR MIT */ +/* +* C Run Time support for NOLIBC +* Copyright (C) 2023 Zhangjin Wu <[email protected]> +*/ + +#ifndef _NOLIBC_CRT_H
Quoted 9 lines
tools/include/nolibc/crt.h line 13
Quoted 6 lines
tools/include/nolibc/crt.h new file +6
-0+#define _NOLIBC_CRT_H + +char **environ __attribute__((weak)); +const unsigned long *_auxv __attribute__((weak)); + +typedef int (_nolibc_main_fn )(int ,char **,char **);What's the advantage of the typedef over using the pointer type inline?
With the extra comment added, this is not required anymore, will remove this typedef.
Quoted 4 lines
Quoted 4 lines
tools/include/nolibc/crt.h new file +4
-0+static void exit (int ); + +void _start_c (long *sp) +{
Quoted 9 lines
tools/include/nolibc/crt.h line 22
Quoted 5 lines
tools/include/nolibc/crt.h new file +5
-0+int argc, i; +char **argv; +char **envp; +/* silence potential warning: conflicting types for 'main' */ +_nolibc_main_fn _nolibc_main __asm__ ("main" );What about the stackprotector initialization?
It would really fit great into this series.
Ok, which gcc version supports stackprotector? seems the test even skip on gcc 10, I will find one to verify the code change.
Quoted 15 lines
Quoted 15 lines
tools/include/nolibc/crt.h new file +15
-0+ +/* +* sp : argc <-- argument count, required by main() +* argv: argv[0] <-- argument vector, required by main() +* argv[1] +* ... +* argv[argc-1] +* null +* envp: envp[0] <-- environment variables, required by main() and getenv() +* envp[1] +* ... +* null +* _auxv: auxv[0] <-- auxiliary vector, required by getauxval() +* auxv[1] +* ...
Quoted 9 lines
tools/include/nolibc/crt.h line 43
Quoted 6 lines
tools/include/nolibc/crt.h new file +6
-0+* null +*/ + +/* assign argc and argv */ + argc = sp[0 ]; + argv = (void *)(sp +1 );Bit of a weird mismatch between array syntax and pointer arithmetic.
This 'argc = *sp;' may be better ;-)
Quoted 7 lines
tools/include/nolibc/crt.h line 47
Quoted 4 lines
tools/include/nolibc/crt.h new file +4
-0+ +/* find envp */ + envp = argv + argc +1 ; + environ = envp;Is envp really needed? Could just be assigned directly to environ.
Ok, let's save one variable, envp is used to be consistent with the frequenty declaration of main().
Quoted 9 lines
tools/include/nolibc/crt.h line 53
Quoted 6 lines
tools/include/nolibc/crt.h new file +6
-0+ +/* find auxv */ + i =0 ; +while (envp[i]) + i++; + _auxv = (void *)(envp + i +1 );Could be simplified a bit:
Quoted 4 lines
_auxv= (void *) envp;
while (_auxv)_auxv++;
Yeah, it is better, but needs a little change.
_auxv = (void *) envp;
while (*_auxv)
_auxv++;
_auxv++;Thanks,
Zhangjin
Quoted 9 lines
Quoted 9 lines
tools/include/nolibc/crt.h new file +6
-0+ +/* go to application */ + exit(_nolibc_main(argc, argv, envp)); +} + +#endif /* _NOLIBC_CRT_H */ Signature · 1 line
2.25.1
Re: [PATCH v3 02/11] tools/nolibc: add new crt.h with _start_c
Hi, Thomas
Quoted 16 lines
Quoted 1 line · Zhangjin Wu
[...]
Quoted 4 lines
Quoted 4 lines
tools/include/nolibc/crt.h new file +4
-0+static void exit (int ); + +void _start_c (long *sp) +{Quoted 9 lines
tools/include/nolibc/crt.h line 22
Quoted 5 lines
tools/include/nolibc/crt.h new file +5
-0+int argc, i; +char **argv; +char **envp; +/* silence potential warning: conflicting types for 'main' */ +_nolibc_main_fn _nolibc_main __asm__ ("main" );What about the stackprotector initialization?
It would really fit great into this series.
Quoted 3 lines
Ok, which gcc version supports stackprotector? seems the test even skip on gcc 10, I will find one to verify the code change.
Thomas, please ignore this question, I forgot some options in my own script.
Best regards,
Zhangjin
Quoted 24 lines
Quoted 24 lines
Quoted 11 lines
tools/include/nolibc/crt.h new file +11
-0+ +/* +* sp : argc <-- argument count, required by main() +* argv: argv[0] <-- argument vector, required by main() +* argv[1] +* ... +* argv[argc-1] +* null +* envp: envp[0] <-- environment variables, required by main() and getenv() +* envp[1] +* ... tools/include/nolibc/crt.h line 43
Quoted 10 lines
tools/include/nolibc/crt.h new file +10
-0+* null +* _auxv: auxv[0] <-- auxiliary vector, required by getauxval() +* auxv[1] +* ... +* null +*/ + +/* assign argc and argv */ + argc = sp[0 ]; + argv = (void *)(sp +1 );Bit of a weird mismatch between array syntax and pointer arithmetic.
Re: [PATCH v3 02/11] tools/nolibc: add new crt.h with _start_c
In reply to Zhangjin WuQuoted 44 lines · Zhangjin Wu
Quoted 1 line · Zhangjin Wu
[..]
Quoted 13 lines
Quoted 13 lines
tools/include/nolibc/crt.h new file +7
-0@@ -0,0 +1,59 @@ +/* SPDX-License-Identifier: LGPL-2.1 OR MIT */ +/* +* C Run Time support for NOLIBC +* Copyright (C) 2023 Zhangjin Wu <[email protected]> +*/ + +#ifndef _NOLIBC_CRT_HQuoted 9 lines
tools/include/nolibc/crt.h line 13
Quoted 6 lines
tools/include/nolibc/crt.h new file +6
-0+#define _NOLIBC_CRT_H + +char **environ __attribute__((weak)); +const unsigned long *_auxv __attribute__((weak)); + +typedef int (_nolibc_main_fn )(int ,char **,char **);What's the advantage of the typedef over using the pointer type inline?
With the extra comment added, this is not required anymore, will remove this typedef.
Quoted 4 lines
Quoted 4 lines
tools/include/nolibc/crt.h new file +4
-0+static void exit (int ); + +void _start_c (long *sp) +{Quoted 9 lines
tools/include/nolibc/crt.h line 22
Quoted 5 lines
tools/include/nolibc/crt.h new file +5
-0+int argc, i; +char **argv; +char **envp; +/* silence potential warning: conflicting types for 'main' */ +_nolibc_main_fn _nolibc_main __asm__ ("main" );What about the stackprotector initialization?
It would really fit great into this series.
Quoted 2 lines
Ok, which gcc version supports stackprotector? seems the test even skip on gcc 10, I will find one to verify the code change.
For crosscompilation I use the crosstools from kernel.org at
https://mirrors.edge.kernel.org/pub/tools/crosstool/
It seems to be at least in their gcc 9.5.0.
And also in their 11.2 which I use mostly at the moment.
Quoted 51 lines
Quoted 14 lines
Quoted 14 lines
tools/include/nolibc/crt.h new file +14
-0+/* +* sp : argc <-- argument count, required by main() +* argv: argv[0] <-- argument vector, required by main() +* argv[1] +* ... +* argv[argc-1] +* null +* envp: envp[0] <-- environment variables, required by main() and getenv() +* envp[1] +* ... +* null +* _auxv: auxv[0] <-- auxiliary vector, required by getauxval() +* auxv[1] +* ... Quoted 9 lines
tools/include/nolibc/crt.h line 43
Quoted 6 lines
tools/include/nolibc/crt.h new file +6
-0+* null +*/ + +/* assign argc and argv */ + argc = sp[0 ]; + argv = (void *)(sp +1 );Bit of a weird mismatch between array syntax and pointer arithmetic.
This 'argc = *sp;' may be better ;-)
Quoted 7 lines
tools/include/nolibc/crt.h line 47
Quoted 4 lines
tools/include/nolibc/crt.h new file +4
-0+ +/* find envp */ + envp = argv + argc +1 ; + environ = envp;Is envp really needed? Could just be assigned directly to environ.
Ok, let's save one variable, envp is used to be consistent with the frequenty declaration of
main().Quoted 9 lines
tools/include/nolibc/crt.h line 53
Quoted 6 lines
tools/include/nolibc/crt.h new file +6
-0+ +/* find auxv */ + i =0 ; +while (envp[i]) + i++; + _auxv = (void *)(envp + i +1 );Could be simplified a bit:
Quoted 4 lines
_auxv= (void *) envp;
while (_auxv)_auxv++;
Quoted 6 lines
Yeah, it is better, but needs a little change.
_auxv = (void *) envp; while (*_auxv) _auxv++; _auxv++;
Good catch, thanks.
Quoted 1 line
[..]
Thomas
Willy Tarreau
maintainerRe: [PATCH v3 02/11] tools/nolibc: add new crt.h with _start_c
In reply to Zhangjin WuHi Zhangjin,
I haven't reviewed the rest yet but regarding this point:
Quoted 13 lines · Zhangjin Wu
Quoted 8 lines
tools/include/nolibc/crt.h line 53
Quoted 5 lines
tools/include/nolibc/crt.h new file +5
-0+/* find auxv */ + i =0 ; +while (envp[i]) + i++; + _auxv = (void *)(envp + i +1 );Could be simplified a bit:
Quoted 4 lines
_auxv= (void *) envp;
while (_auxv)_auxv++;
Quoted 6 lines
Yeah, it is better, but needs a little change.
_auxv = (void *) envp; while (*_auxv) _auxv++; _auxv++;
Or just:
_auxv = (void*)environ;
while (*_auxv++)
;or:
for (_auxv = (void*)environ; *_auxv++; )
;Please also have a look at the output code, because at low optimization levels, compilers sometimes produce a better code with a local variable than with a global variable in a loop. Thus I wouldn't be that much surprised if at -O0 or -O1 you'd see slightly more compact code using:
/* find envp */
environ = argv + argc + 1;
/* find auxv */
for (auxv = environ; *auxv++)
;
_auxv = auxv;than:
/* find envp */
envp = argv + argc + 1;
environ = envp;
/* find auxv */
for (_auxv = environ; *_auxv++)
;Since it's going to become generic code, it's worth running a few tests to see how to best polish it.
Thanks,
Willy
Re: [PATCH v3 02/11] tools/nolibc: add new crt.h with _start_c
Hi, Willy, Thomas
Quoted 33 lines
Hi Zhangjin,
I haven't reviewed the rest yet but regarding this point:
Quoted 13 lines · Zhangjin Wu
Quoted 8 lines
tools/include/nolibc/crt.h line 53
Quoted 5 lines
tools/include/nolibc/crt.h new file +5
-0+/* find auxv */ + i =0 ; +while (envp[i]) + i++; + _auxv = (void *)(envp + i +1 );Could be simplified a bit:
Quoted 4 lines
_auxv= (void *) envp;
while (_auxv)_auxv++;Quoted 6 lines
Yeah, it is better, but needs a little change.
_auxv = (void *) envp; while (*_auxv) _auxv++; _auxv++;Or just:
_auxv = (void*)environ; while (*_auxv++) ;or:
Quoted 8 lines
for (_auxv = (void*)environ; *_auxv++; ) ;Please also have a look at the output code, because at low optimization levels, compilers sometimes produce a better code with a local variable than with a global variable in a loop. Thus I wouldn't be that much surprised if at -O0 or -O1 you'd see slightly more compact code using:
Very good suggestion, I did find some interesting results, after removing some local variables, although the text size shrinks, but the total size is the same with -O0/1/2/3/s.
Here is the diff with -O0/1/2/3/s (a REPORT_SIZE macro may be required for this):
$ diff -Nubr startup.old.txt startup.new.txt
--- startup.old.txt 2023-07-14 10:22:45.990413661 +0800
+++ startup.new.txt 2023-07-14 10:22:52.278869146 +0800
@@ -2,34 +2,34 @@
sudo strip -s nolibc-test
size nolibc-test
text data bss dec hex filename
- 46872 88 80 47040 b7c0 nolibc-test
+ 46836 88 80 47004 b79c nolibc-test
wc -c nolibc-test
58016 nolibc-test
O1:
sudo strip -s nolibc-test
size nolibc-test
text data bss dec hex filename
- 27298 88 72 27458 6b42 nolibc-test
+ 27287 88 72 27447 6b37 nolibc-test
wc -c nolibc-test
37536 nolibc-test
O2:
sudo strip -s nolibc-test
size nolibc-test
text data bss dec hex filename
- 25688 88 72 25848 64f8 nolibc-test
+ 25672 88 72 25832 64e8 nolibc-test
wc -c nolibc-test
37536 nolibc-test
O3:
sudo strip -s nolibc-test
size nolibc-test
text data bss dec hex filename
- 44020 88 72 44180 ac94 nolibc-test
+ 44004 88 72 44164 ac84 nolibc-test
wc -c nolibc-test
53920 nolibc-test
Os:
sudo strip -s nolibc-test
size nolibc-test
text data bss dec hex filename
- 20887 88 72 21047 5237 nolibc-test
+ 20884 88 72 21044 5234 nolibc-test
wc -c nolibc-test
33440 nolibc-testThe code with local variables:
void _start_c(long *sp)
{
int argc, i;
char **argv;
char **envp;
/* silence potential warning: conflicting types for 'main' */
int _nolibc_main(int, char **, char **) __asm__ ("main");
/*
* sp : argc <-- argument count, required by main()
* argv: argv[0] <-- argument vector, required by main()
* argv[1]
* ...
* argv[argc-1]
* null
* environ: environ[0] <-- environment variables, required by main() and getenv()
* environ[1]
* ...
* null
* _auxv: _auxv[0] <-- auxiliary vector, required by getauxval()
* _auxv[1]
* ...
* null
*/
/* assign argc and argv */
argc = *sp;
argv = (void *)(sp + 1);
/* find environ */
environ = envp = argv + argc + 1;
/* find _auxv */
for (i = 0; *(envp + i); i++)
;
_auxv = (void *)(envp + i + 1);
/* go to application */
exit(_nolibc_main(argc, argv, envp));
}The code with global variables:
void _start_c(long *sp)
{
int argc;
char **argv;
/* silence potential warning: conflicting types for 'main' */
int _nolibc_main(int, char **, char **) __asm__ ("main");
/*
* sp : argc <-- argument count, required by main()
* argv: argv[0] <-- argument vector, required by main()
* argv[1]
* ...
* argv[argc-1]
* null
* environ: environ[0] <-- environment variables, required by main() and getenv()
* environ[1]
* ...
* null
* _auxv: _auxv[0] <-- auxiliary vector, required by getauxval()
* _auxv[1]
* ...
* null
*/
/* assign argc and argv */
argc = *sp;
argv = (void *)(sp + 1);
/* find environ */
environ = argv + argc + 1;
/* find _auxv */
for (_auxv = (void *)environ; *_auxv++;)
;
/* go to application */
exit(_nolibc_main(argc, argv, environ));
}Which one do you prefer? the one with local variables may be more readable (not that much), the one with global variables has smaller text size (and therefore smaller memory footprint).
BTW, just found an arch-<ARCH>.h bug with -O0, seems the 'optimize("omit-frame-pointer")' attribute not really work as expected with -O0. It uses frame pointer for _start eventually and breaks the stack pointer variable (a push 'rbp' inserted at the begging of _start, so, the real rsp has an offset), so, therefore, it is not able to get the right argc, argv, environ and _auxv with -O0 currently. A solution is reverting _start to pure assembly.
For the above tests, I manually reverted the arch-x86_64.h to:
__asm__(
".text\n"
".weak _start\n"
"_start:\n"
#ifdef _NOLIBC_STACKPROTECTOR
"call __stack_chk_init\n" /* initialize stack protector */
#endif
"xor %ebp, %ebp\n" /* zero the stack frame */
"mov %rsp, %rdi\n" /* save stack pointer to %rdi, as arg1 of _start_c */
"and $-16, %rsp\n" /* %rsp must be 16-byte aligned before call */
"call _start_c\n" /* transfer to c runtime */
"hlt\n" /* ensure it does not return */
);'man gcc' shows:
Most optimizations are completely disabled at -O0 or if an -O level is not set on the command line, even if individual optimization flags are specified.
To want -O0 work again, since now we have C _start_c, is it ok for us to revert the commit 7f8548589661 ("tools/nolibc: make compiler and assembler agree on the section around _start") and the later __no_stack_protector changes?
At the same time, to verify such issues, as Thomas suggested, in this series, we may need to add more startup tests to verify argc, argv, environ, _auxv, do we need to add a standalone run_startup (or run_crt) test entry just like run_syscall? or, let's simply add more in the run_stdlib, like the environ test added by Thomas. seems, the argc test is necessary one currently missing (argc
Quoted 1 line
= 1):
CASE_TEST(argc); EXPECT_GE(1, test_argc, 1); break;
As we summarized before, the related test cases are:
argv0:
CASE_TEST(chmod_argv0); EXPECT_SYSZR(1, chmod(test_argv0, 0555)); break;
CASE_TEST(chroot_exe); EXPECT_SYSER(1, chroot(test_argv0), -1, ENOTDIR); break;environ:
CASE_TEST(chdir_root); EXPECT_SYSZR(1, chdir("/")); chdir(getenv("PWD")); break;
CASE_TEST(environ); EXPECT_PTREQ(1, environ, test_envp); break;
CASE_TEST(getenv_TERM); EXPECT_STRNZ(1, getenv("TERM")); break;
CASE_TEST(getenv_blah); EXPECT_STRZR(1, getenv("blah")); break;auxv:
CASE_TEST(getpagesize); EXPECT_SYSZR(1, test_getpagesize()); break;
The above tests are in different test group and are not aimed to startup test, we'd better add a run_startup (or run_crt) test group before any other tests, it is a requiremnt of the others, we at least have these ones:
+int run_startup(int min, int max)
+{
+ int test;
+ int tmp;
+ int ret = 0;
+
+ for (test = min; test >= 0 && test <= max; test++) {
+ int llen = 0; /* line length */
+
+ /* avoid leaving empty lines below, this will insert holes into
+ * test numbers.
+ */
+ switch (test + __LINE__ + 1) {
+ CASE_TEST(argc); EXPECT_GE(1, test_argc, 1); break;
+ CASE_TEST(argv_addr); EXPECT_PTRNZ(1, test_argv); break;
+ CASE_TEST(argv_total); EXPECT_EQ(1, environ - test_argv - 1, test_argc); break;
+ CASE_TEST(argv0_addr); EXPECT_PTRNZ(1, argv0); break;
+ CASE_TEST(argv0_str); EXPECT_STRNZ(1, argv0); break;
+ CASE_TEST(argv0_len); EXPECT_GE(1, strlen(argv0), 1); break;
+ CASE_TEST(environ_addr); EXPECT_PTRNZ(1, environ); break;
+ CASE_TEST(environ_envp); EXPECT_PTREQ(1, environ, test_envp); break;
+ CASE_TEST(environ_total); EXPECT_GE(1, (void *)_auxv - (void *)environ - 1, 1); break;
+ CASE_TEST(_auxv_addr); EXPECT_PTRNZ(1, _auxv); break;
+ case __LINE__:
+ return ret; /* must be last */
+ /* note: do not set any defaults so as to permit holes above */
+ }
+ }
+ return ret;
+}
Any more?
Quoted 8 lines
/* find envp */
environ = argv + argc + 1;
/* find auxv */
for (auxv = environ; *auxv++)
;
_auxv = auxv;Quoted 9 lines
than:
/* find envp */ envp = argv + argc + 1; environ = envp; /* find auxv */ for (_auxv = environ; *_auxv++) ;
Great, *_auxv++ is a very good trick to avoid duplicated _auxv++, although it is a little hard to read (not that hard in reality) ;-)
Quoted 3 lines
Since it's going to become generic code, it's worth running a few tests to see how to best polish it.
Yes, I focused on the big shrinking itself but forgot to polish the _start_c itself ;-)
Best regards,
Zhangjin
Quoted 2 lines
Thanks,
Willy
Re: [PATCH v3 02/11] tools/nolibc: add new crt.h with _start_c
Quoted 5 lines · Zhangjin Wu
[..]
Which one do you prefer? the one with local variables may be more readable (not that much), the one with global variables has smaller text size (and therefore smaller memory footprint).
The one with local variables. But not by much.
Quoted 24 lines
BTW, just found an arch-<ARCH>.h bug with -O0, seems the '
optimize("omit-frame-pointer")' attribute not really work as expected with -O0. It uses frame pointer for_starteventually and breaks the stack pointer variable (a push 'rbp' inserted at the begging of_start, so, the real rsp has an offset), so, therefore, it is not able to get the right argc, argv, environ and_auxvwith -O0 currently. A solution is reverting_startto pure assembly.For the above tests, I manually reverted the
arch-x86_64.hto:__asm__( ".text\n" ".weak _start\n" "_start:\n" #ifdef _NOLIBC_STACKPROTECTOR "call __stack_chk_init\n" /* initialize stack protector */ #endif "xor %ebp, %ebp\n" /* zero the stack frame */ "mov %rsp, %rdi\n" /* save stack pointer to %rdi, as arg1 of _start_c */ "and $-16, %rsp\n" /* %rsp must be 16-byte aligned before call */ "call _start_c\n" /* transfer to c runtime */ "hlt\n" /* ensure it does not return */ );
Quoted 7 lines
'man gcc' shows:
Most optimizations are completely disabled at -O0 or if an -O level is not set on the command line, even if individual optimization flags are specified.
To want -O0 work again, since now we have C
_start_c, is it ok for us to revert the commit7f8548589661("tools/nolibc: make compiler and assembler agree on the section around_start") and the later__no_stack_protectorchanges?
This commit explicitly mentions being tested with -O0 on x86_64. I was also not able to reproduce the issue.
Before doing any reverts I think some more investigation is in order. Can you provide exact reproduction steps?
Quoted 62 lines
At the same time, to verify such issues, as Thomas suggested, in this series, we may need to add more startup tests to verify argc, argv, environ,
_auxv, do we need to add a standalonerun_startup(orrun_crt) test entry just likerun_syscall? or, let's simply add more in therun_stdlib, like the environ test added by Thomas. seems, the argc test is necessary one currently missing (argcQuoted 1 line
= 1):
CASE_TEST(argc); EXPECT_GE(1, test_argc, 1); break;As we summarized before, the related test cases are:
argv0:
CASE_TEST(chmod_argv0); EXPECT_SYSZR(1, chmod(test_argv0, 0555)); break; CASE_TEST(chroot_exe); EXPECT_SYSER(1, chroot(test_argv0), -1, ENOTDIR); break;environ:
CASE_TEST(chdir_root); EXPECT_SYSZR(1, chdir("/")); chdir(getenv("PWD")); break; CASE_TEST(environ); EXPECT_PTREQ(1, environ, test_envp); break; CASE_TEST(getenv_TERM); EXPECT_STRNZ(1, getenv("TERM")); break; CASE_TEST(getenv_blah); EXPECT_STRZR(1, getenv("blah")); break;auxv:
CASE_TEST(getpagesize); EXPECT_SYSZR(1, test_getpagesize()); break;The above tests are in different test group and are not aimed to startup test, we'd better add a
run_startup(orrun_crt) test group before any other tests, it is a requiremnt of the others, we at least have these ones:+int
run_startup(int min, int max)
+{
+ int test;
+ int tmp;
+ int ret = 0;
+
+ for (test = min; test >= 0 && test <= max; test++) {
+ int llen = 0; /* line length */
+
+ /* avoid leaving empty lines below, this will insert holes into
+ * test numbers.
+ */
+ switch (test +__LINE__+ 1) {
+CASE_TEST(argc);EXPECT_GE(1, test_argc, 1); break;
+CASE_TEST(argv_addr);EXPECT_PTRNZ(1, test_argv); break;
+CASE_TEST(argv_total);EXPECT_EQ(1, environ - test_argv - 1, test_argc); break;
+CASE_TEST(argv0_addr);EXPECT_PTRNZ(1, argv0); break;
+CASE_TEST(argv0_str);EXPECT_STRNZ(1, argv0); break;
+CASE_TEST(argv0_len);EXPECT_GE(1, strlen(argv0), 1); break;
+CASE_TEST(environ_addr);EXPECT_PTRNZ(1, environ); break;
+CASE_TEST(environ_envp);EXPECT_PTREQ(1, environ, test_envp); break;
+CASE_TEST(environ_total);EXPECT_GE(1, (void *)_auxv - (void *)environ - 1, 1); break;
+CASE_TEST(_auxv_addr);EXPECT_PTRNZ(1, _auxv); break;
+ case__LINE__:
+ return ret; /* must be last */
+ /* note: do not set any defaults so as to permit holes above */
+ }
+ }
+ return ret;
+}
Quoted 1 line
Any more?
My original idea was to have tests that exec /proc/self/exe or argv0. This way we can actually pass and validate arbitrary argc, argv and environ values.
But looking at your list, that should be enough.
Quoted 1 line
[..]
Re: [PATCH v3 02/11] tools/nolibc: add new crt.h with _start_c
Quoted 36 lines · Zhangjin Wu
Quoted 6 lines
[..]
Which one do you prefer? the one with local variables may be more readable (not that much), the one with global variables has smaller text size (and therefore smaller memory footprint).
The one with local variables. But not by much.
Quoted 26 lines
BTW, just found an arch-<ARCH>.h bug with -O0, seems the '
optimize("omit-frame-pointer")' attribute not really work as expected with -O0. It uses frame pointer for_starteventually and breaks the stack pointer variable (a push 'rbp' inserted at the begging of_start, so, the real rsp has an offset), so, therefore, it is not able to get the right argc, argv, environ and_auxvwith -O0 currently. A solution is reverting_startto pure assembly.For the above tests, I manually reverted the
arch-x86_64.hto:__asm__( ".text\n" ".weak _start\n" "_start:\n" #ifdef _NOLIBC_STACKPROTECTOR "call __stack_chk_init\n" /* initialize stack protector */ #endif "xor %ebp, %ebp\n" /* zero the stack frame */ "mov %rsp, %rdi\n" /* save stack pointer to %rdi, as arg1 of _start_c */ "and $-16, %rsp\n" /* %rsp must be 16-byte aligned before call */ "call _start_c\n" /* transfer to c runtime */ "hlt\n" /* ensure it does not return */ );'man gcc' shows:
Quoted 7 lines
Quoted 5 lines
Most optimizations are completely disabled at -O0 or if an -O level is not set on the command line, even if individual optimization flags are specified.
To want -O0 work again, since now we have C
_start_c, is it ok for us to revert the commit7f8548589661("tools/nolibc: make compiler and assembler agree on the section around_start") and the later__no_stack_protectorchanges?This commit explicitly mentions being tested with -O0 on
x86_64.
Yeah, I was worried about that the old tests didn't use any of the startup variables, but the getpagesize test may be a very old test, It uses _auxv and should fail If -O0 not work.
Quoted 2 lines
I was also not able to reproduce the issue.
Thanks very much for your 'reproduce' result, It is so weird, just rechecked the toolchain, 13.1.0 from https://mirrors.edge.kernel.org/ is ok, gcc 9, gcc 10.3 not work.
But even in the page of 13.1.0 [1], we still see this line:
Most optimizations are completely disabled at -O0 or if an -O level is not set on the command line, even if individual optimization flags are specified.
Not sure if "individual optimization flags" also means the optimize() flags in gcc attributes. or the doc is not updated yet?
And further found gcc 11.1.0 is ok, gcc 10.4 still not work, so, gcc 11.1.0 may changed something to let the "individual optimization flags" work with -O0.
We may need to at least document this issue in some files, -O0 is not such a frequently-used option, not sure if we still need -O0 work with the older gcc < 11.1.0 ;-)
Willy, I'm not sure if the issues solved by the commit 7f8548589661 ("tools/nolibc: make compiler and assembler agree on the section around _start") still exist after we using _start_c()?
Thomas, because we plan to move the stackprotector init to _start_c(), If using pure assembly _start, we may also not need the __no_stack_protector macro too?
Welcome more discussion.
[1]: https://gcc.gnu.org/onlinedocs/gcc-13.1.0/gcc/Optimize-Options.html
Quoted 3 lines
Before doing any reverts I think some more investigation is in order. Can you provide exact reproduction steps?
some startup variables related tests failed with gcc 9 and gcc 10 (even with gcc 10.4.0):
$ x86_64-linux-gcc --version
x86_64-linux-gcc (GCC) 10.4.0
Copyright (C) 2020 Free Software Foundation, Inc.
This is free software; see the source for copying conditions. There is NO
warranty; not even for MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. $ make run-user CROSS_COMPILE=x86_64-linux-
0 argc = 0 [FAIL]
3 argv_total = 0 [FAIL]
4 argv0_addr = <0x1> [FAIL]
5 argv0_str = <(null)> [FAIL]
6 argv0_len = 0 [FAIL]
14 chmod_argv0 = -1 EFAULT [FAIL]
19 chroot_exe = -1 EFAULT != (-1 ENOTDIR) [FAIL]
0 getenv_TERM = <(null)> [FAIL]Quoted 62 lines
Quoted 62 lines
At the same time, to verify such issues, as Thomas suggested, in this series, we may need to add more startup tests to verify argc, argv, environ,
_auxv, do we need to add a standalonerun_startup(orrun_crt) test entry just likerun_syscall? or, let's simply add more in therun_stdlib, like the environ test added by Thomas. seems, the argc test is necessary one currently missing (argcQuoted 1 line
= 1):
CASE_TEST(argc); EXPECT_GE(1, test_argc, 1); break;As we summarized before, the related test cases are:
argv0:
CASE_TEST(chmod_argv0); EXPECT_SYSZR(1, chmod(test_argv0, 0555)); break; CASE_TEST(chroot_exe); EXPECT_SYSER(1, chroot(test_argv0), -1, ENOTDIR); break;environ:
CASE_TEST(chdir_root); EXPECT_SYSZR(1, chdir("/")); chdir(getenv("PWD")); break; CASE_TEST(environ); EXPECT_PTREQ(1, environ, test_envp); break; CASE_TEST(getenv_TERM); EXPECT_STRNZ(1, getenv("TERM")); break; CASE_TEST(getenv_blah); EXPECT_STRZR(1, getenv("blah")); break;auxv:
CASE_TEST(getpagesize); EXPECT_SYSZR(1, test_getpagesize()); break;The above tests are in different test group and are not aimed to startup test, we'd better add a
run_startup(orrun_crt) test group before any other tests, it is a requiremnt of the others, we at least have these ones:+int
run_startup(int min, int max)
+{
+ int test;
+ int tmp;
+ int ret = 0;
+
+ for (test = min; test >= 0 && test <= max; test++) {
+ int llen = 0; /* line length */
+
+ /* avoid leaving empty lines below, this will insert holes into
+ * test numbers.
+ */
+ switch (test +__LINE__+ 1) {
+CASE_TEST(argc);EXPECT_GE(1, test_argc, 1); break;
+CASE_TEST(argv_addr);EXPECT_PTRNZ(1, test_argv); break;
+CASE_TEST(argv_total);EXPECT_EQ(1, environ - test_argv - 1, test_argc); break;
+CASE_TEST(argv0_addr);EXPECT_PTRNZ(1, argv0); break;
+CASE_TEST(argv0_str);EXPECT_STRNZ(1, argv0); break;
+CASE_TEST(argv0_len);EXPECT_GE(1, strlen(argv0), 1); break;
+CASE_TEST(environ_addr);EXPECT_PTRNZ(1, environ); break;
+CASE_TEST(environ_envp);EXPECT_PTREQ(1, environ, test_envp); break;
+CASE_TEST(environ_total);EXPECT_GE(1, (void *)_auxv - (void *)environ - 1, 1); break;
+CASE_TEST(_auxv_addr);EXPECT_PTRNZ(1, _auxv); break;
+ case__LINE__:
+ return ret; /* must be last */
+ /* note: do not set any defaults so as to permit holes above */
+ }
+ }
+ return ret;
+}
Quoted 8 lines
Quoted 1 line
Any more?
My original idea was to have tests that exec
/proc/self/exeor argv0. This way we can actually pass and validate arbitrary argc, argv and environ values.But looking at your list, that should be enough.
Ok.
Best regards,
Zhangjin
Quoted 1 line
Quoted 1 line
[..]
Re: [PATCH v3 02/11] tools/nolibc: add new crt.h with _start_c
Quoted 18 lines · Zhangjin Wu
Quoted 1 line · Zhangjin Wu
[..]
Quoted 2 lines
I was also not able to reproduce the issue.
Thanks very much for your 'reproduce' result, It is so weird, just rechecked the toolchain, 13.1.0 from https://mirrors.edge.kernel.org/ is ok, gcc 9, gcc 10.3 not work.
But even in the page of 13.1.0 [1], we still see this line:
Most optimizations are completely disabled at -O0 or if an -O level is not set on the command line, even if individual optimization flags are specified.
Not sure if "individual optimization flags" also means the
optimize()flags in gcc attributes. or the doc is not updated yet?
Quoted 7 lines
And further found gcc 11.1.0 is ok, gcc 10.4 still not work, so, gcc 11.1.0 may changed something to let the "individual optimization flags" work with -O0.
We may need to at least document this issue in some files, -O0 is not such a frequently-used option, not sure if we still need -O0 work with the older gcc < 11.1.0 ;-)
It seems we can avoid the issue by enforcing optimizations for _start:
tools/include/nolibc/arch-x86_64.h +1 -1
@@ -161,12 +161,9 @@* 2) The deepest stack frame should be zero (the %rbp). * */ -void __attribute__((weak, noreturn, optimize("omit-frame-pointer"))) __no_stack_protector _start(void) +void __attribute__((weak, noreturn, optimize("Os", "omit-frame-pointer"))) __no_stack_protector _start(void)
Quoted 7 lines
Willy, I'm not sure if the issues solved by the commit
7f8548589661("tools/nolibc: make compiler and assembler agree on the section around_start") still exist after we using_start_c()?Thomas, because we plan to move the stackprotector init to
_start_c(), If using pure assembly_start, we may also not need the__no_stack_protectormacro too?
It would probably not needed anymore in this case.
Thomas
Willy Tarreau
maintainerRe: [PATCH v3 01/11] tools/nolibc: remove the old sys_stat support
In reply to Zhangjin WuHi Zhangjin,
Quoted 11 lines · Zhangjin Wu
__NR_statxhas been added from v4.10:commit
a528d35e8bfc("statx: Add a system call to make enhanced file info available")It has been supported by all of the platforms since at least from v4.20 and glibc 2.28.
Let's remove the old arch related and dependent
sys_statsupport completely.This is friendly to the future new architecture porting.
As I previously said, I'd like that we at least preserve compatibility with supported stable branches. 4.14 and 4.19 are still supported, so does this mean that if I rebuild my preinit against the updated nolibc it will fail to boot on such older systems for the archs that we support?
Because if it means that in order to support all currently active kernels, one must only build from an older copy of the lib, that becomes a disservice to its development and usage. Thus if you checked that aarch64, arm, i386, mips, riscv, s390 and x86_64 had already adopted statx by 4.14, then I'm fine and we can drop stat(), but then it must be mentioned in the commit message, because here it's not explicit.
Thanks!
Willy
Re: [PATCH v3 02/11] tools/nolibc: add new crt.h with _start_c
In reply to Thomas WeißschuhQuoted 34 lines · Zhangjin Wu
Quoted 18 lines
Quoted 1 line · Zhangjin Wu
[..]
Quoted 2 lines
I was also not able to reproduce the issue.
Thanks very much for your 'reproduce' result, It is so weird, just rechecked the toolchain, 13.1.0 from https://mirrors.edge.kernel.org/ is ok, gcc 9, gcc 10.3 not work.
But even in the page of 13.1.0 [1], we still see this line:
Most optimizations are completely disabled at -O0 or if an -O level is not set on the command line, even if individual optimization flags are specified.
Not sure if "individual optimization flags" also means the
optimize()flags in gcc attributes. or the doc is not updated yet?Quoted 7 lines
And further found gcc 11.1.0 is ok, gcc 10.4 still not work, so, gcc 11.1.0 may changed something to let the "individual optimization flags" work with -O0.
We may need to at least document this issue in some files, -O0 is not such a frequently-used option, not sure if we still need -O0 work with the older gcc < 11.1.0 ;-)
It seems we can avoid the issue by enforcing optimizations for
_start:tools/include/nolibc/arch-x86_64.h +0
-0
tools/include/nolibc/arch-x86_64.h line 165
Quoted 6 lines
tools/include/nolibc/arch-x86_64.h +1
-1@@ -161,12 +161,9 @@* 2) The deepest stack frame should be zero (the %rbp). * */ -void __attribute__((weak, noreturn, optimize("omit-frame-pointer"))) __no_stack_protector _start(void) +void __attribute__((weak, noreturn, optimize("Os", "omit-frame-pointer"))) __no_stack_protector _start(void)
Great, it works and it is minimal enough ;-)
Thanks very much.
Quoted 10 lines
Quoted 7 lines
Willy, I'm not sure if the issues solved by the commit
7f8548589661("tools/nolibc: make compiler and assembler agree on the section around_start") still exist after we using_start_c()?Thomas, because we plan to move the stackprotector init to
_start_c(), If using pure assembly_start, we may also not need the__no_stack_protectormacro too?It would probably not needed anymore in this case.
Yeah, but let's reserve it as-is for we have the working omit-frame-pointer now.
Best regards,
Zhangjin
Quoted 1 line
Thomas
Willy Tarreau
maintainerRe: [PATCH v3 02/11] tools/nolibc: add new crt.h with _start_c
In reply to Zhangjin WuHi Zhangjin,
Quoted 18 lines · Zhangjin Wu
Quoted 10 lines
or:
for (_auxv = (void*)environ; *_auxv++; ) ;Please also have a look at the output code, because at low optimization levels, compilers sometimes produce a better code with a local variable than with a global variable in a loop. Thus I wouldn't be that much surprised if at -O0 or -O1 you'd see slightly more compact code using:
Very good suggestion, I did find some interesting results, after removing some local variables, although the text size shrinks, but the total size is the same with -O0/1/2/3/s.
Here is the diff with -O0/1/2/3/s (a
REPORT_SIZEmacro may be required for this):
Quoted 9 lines
$ diff -Nubr startup.old.txt startup.new.txt
--- startup.old.txt 2023-07-14 10:22:45.990413661 +0800
+++ startup.new.txt 2023-07-14 10:22:52.278869146 +0800
@@ -2,34 +2,34 @@
sudo strip -s nolibc-test
size nolibc-test
text data bss dec hex filename
- 46872 88 80 47040 b7c0 nolibc-test
+ 46836 88 80 47004 b79c nolibc-test(...)
I meant only checking the function's size, not the whole program :-)
Here's what I'm having for the function:
$ for opt in O0 O1 O2 Os O3; do
echo "## $opt"
for file in global local local2; do
gcc -$opt -c startc-$file.c
done
nm -o --size startc-*.o | grep _start_c
echo
done ## O0
startc-global.o:0000000000000089 T _start_c
startc-local.o:00000000000000ad T _start_c
startc-local2.o:0000000000000090 T _start_c
## O1
startc-global.o:0000000000000048 T _start_c
startc-local.o:0000000000000054 T _start_c
startc-local2.o:000000000000003a T _start_c
## O2
startc-global.o:0000000000000044 T _start_c
startc-local.o:000000000000004d T _start_c
startc-local2.o:0000000000000040 T _start_c
## Os
startc-global.o:0000000000000041 T _start_c
startc-local.o:0000000000000040 T _start_c
startc-local2.o:000000000000003a T _start_c
## O3
startc-global.o:0000000000000044 T _start_c
startc-local.o:000000000000004d T _start_c
startc-local2.o:0000000000000040 T _start_c
In local2 that's always smaller I've just used a local auxv variable instead of iterating over the global one, and it's cheaper than using an index since 1) the instructions are shorter, and 2) the value is already available, no need to initialize one register to zero:
void _start_c(long *sp)
{
long argc;
char **argv;
char **envp;
const unsigned long *auxv;
/* silence potential warning: conflicting types for 'main' */
int _nolibc_main(int, char **, char **) __asm__ ("main");
/*
* sp : argc <-- argument count, required by main()
* argv: argv[0] <-- argument vector, required by main()
* argv[1]
* ...
* argv[argc-1]
* null
* environ: environ[0] <-- environment variables, required by main() and getenv()
* environ[1]
* ...
* null
* _auxv: _auxv[0] <-- auxiliary vector, required by getauxval()
* _auxv[1]
* ...
* null
*/
/* assign argc and argv */
argc = *sp;
argv = (void *)(sp + 1);
/* find environ */
environ = envp = argv + argc + 1;
/* find _auxv */
for (auxv = (void*)envp; *auxv++; )
;
_auxv = auxv;
/* go to application */
exit(_nolibc_main(argc, argv, envp));
}I'm not showing how ugly it is at -O1, as I suspected the global version does indeed perform a write to _auxv for each turn. At -Os it's interesting:
In the global version it looks like this (rdx has envp):
13: 48 8d 42 08 lea 0x8(%rdx),%rax 17: 48 89 15 00 00 00 00 mov %rdx,0x0(%rip) # environ 1e: 48 89 c7 mov %rax,%rdi 21: 48 83 c0 08 add $0x8,%rax 25: 48 83 78 f0 00 cmpq $0x0,-0x10(%rax) 2a: 75 f2 jne 1e <_start_c+0x1e>
In your local version with the index, it looks like this (rdx has envp):
13: 31 c0 xor %eax,%eax 15: 48 89 15 00 00 00 00 mov %rdx,0x0(%rip) # environ 1c: 48 ff c0 inc %rax 1f: 48 83 7c c2 f8 00 cmpq $0x0,-0x8(%rdx,%rax,8) 25: 75 f5 jne 1c <_start_c+0x1c> 27: 48 8d 04 c2 lea (%rdx,%rax,8),%rax # rax now has auxv
In the one without index it's like this:
13: 48 89 15 00 00 00 00 mov %rdx,0x0(%rip) # environ 1a: 48 89 d0 mov %rdx,%rax 1d: 48 83 c0 08 add $0x8,%rax 21: 48 83 78 f8 00 cmpq $0x0,-0x8(%rax) 26: 75 f5 jne 1d <_start_c+0x1d>
Finally, since argc is used in pointer computation while being taken from a long*, it experiences a double conversion while doing so. Storing it as a long avoids this saving 3 extra bytes.
Quoted 3 lines
Which one do you prefer? the one with local variables may be more readable (not that much), the one with global variables has smaller text size (and therefore smaller memory footprint).
The smaller one with local variables and no index ;-)
Quoted 6 lines
BTW, just found an arch-<ARCH>.h bug with -O0, seems the '
optimize("omit-frame-pointer")' attribute not really work as expected with -O0. It uses frame pointer for_starteventually and breaks the stack pointer variable (a push 'rbp' inserted at the begging of_start, so, the real rsp has an offset), so, therefore, it is not able to get the right argc, argv, environ and_auxvwith -O0 currently. A solution is reverting_startto pure assembly.
I didn't notice this one.
Quoted 5 lines
'man gcc' shows:
Most optimizations are completely disabled at -O0 or if an -O level is not set on the command line, even if individual optimization flags are specified.
Indeed, that's pretty clear!
Quoted 3 lines
To want -O0 work again, since now we have C
_start_c, is it ok for us to revert the commit7f8548589661("tools/nolibc: make compiler and assembler agree on the section around_start") and the later__no_stack_protectorchanges?
As Thomas said in somewhere else in the thread, let's analyze deeper first. Maybe passing optimize("O") in addition will be sufficient.
Quoted 2 lines
At the same time, to verify such issues, as Thomas suggested, in this series, we may need to add more startup tests to verify argc, argv, environ,
_auxv, do
Yep!
Quoted 4 lines
we need to add a standalone
run_startup(orrun_crt) test entry just likerun_syscall? or, let's simply add more in therun_stdlib, like the environ test added by Thomas. seems, the argc test is necessary one currently missing (argcQuoted 1 line
= 1):
Yes it could be a good idea to add a startup test category. Some of the stuff we've placed into "stdlib" are more about startup code (provided by the lib) that tends to be very sensitive to the architecture and optimizations.
Quoted 55 lines
CASE_TEST(argc); EXPECT_GE(1, test_argc, 1); break;As we summarized before, the related test cases are:
argv0:
CASE_TEST(chmod_argv0); EXPECT_SYSZR(1, chmod(test_argv0, 0555)); break; CASE_TEST(chroot_exe); EXPECT_SYSER(1, chroot(test_argv0), -1, ENOTDIR); break;environ:
CASE_TEST(chdir_root); EXPECT_SYSZR(1, chdir("/")); chdir(getenv("PWD")); break; CASE_TEST(environ); EXPECT_PTREQ(1, environ, test_envp); break; CASE_TEST(getenv_TERM); EXPECT_STRNZ(1, getenv("TERM")); break; CASE_TEST(getenv_blah); EXPECT_STRZR(1, getenv("blah")); break;auxv:
CASE_TEST(getpagesize); EXPECT_SYSZR(1, test_getpagesize()); break;The above tests are in different test group and are not aimed to startup test, we'd better add a
run_startup(orrun_crt) test group before any other tests, it is a requiremnt of the others, we at least have these ones:+int
run_startup(int min, int max)
+{
+ int test;
+ int tmp;
+ int ret = 0;
+
+ for (test = min; test >= 0 && test <= max; test++) {
+ int llen = 0; /* line length */
+
+ /* avoid leaving empty lines below, this will insert holes into
+ * test numbers.
+ */
+ switch (test +__LINE__+ 1) {
+CASE_TEST(argc);EXPECT_GE(1, test_argc, 1); break;
+CASE_TEST(argv_addr);EXPECT_PTRNZ(1, test_argv); break;
+CASE_TEST(argv_total);EXPECT_EQ(1, environ - test_argv - 1, test_argc); break;
+CASE_TEST(argv0_addr);EXPECT_PTRNZ(1, argv0); break;
+CASE_TEST(argv0_str);EXPECT_STRNZ(1, argv0); break;
+CASE_TEST(argv0_len);EXPECT_GE(1, strlen(argv0), 1); break;
+CASE_TEST(environ_addr);EXPECT_PTRNZ(1, environ); break;
+CASE_TEST(environ_envp);EXPECT_PTREQ(1, environ, test_envp); break;
+CASE_TEST(environ_total);EXPECT_GE(1, (void *)_auxv - (void *)environ - 1, 1); break;
+CASE_TEST(_auxv_addr);EXPECT_PTRNZ(1, _auxv); break;
+ case__LINE__:
+ return ret; /* must be last */
+ /* note: do not set any defaults so as to permit holes above */
+ }
+ }
+ return ret;
+}
Quoted 1 line
Any more?
I quickly glanced over this but I tend to like it, indeed.
Thanks!
Willy
Willy Tarreau
maintainerQuoted 10 lines · Zhangjin Wu
Hi, Willy, Thomas
Here is the revision of the v2 arch support shrink patchset [1], it mainly applies suggestions from you.
It is based on the 20230710-nolibc-ser2-tom-syscall-configv4-report branch of nolibc repo.
Tested for all of the supported archs:
Quoted 13 lines
arch/board | result
------------|------------
arm/versatilepb | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.
arm/vexpress-a9 | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.
arm/virt | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.
aarch64/virt | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.
i386/pc | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.
x86_64/pc | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.
mipsel/malta | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.
loongarch64/virt | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.
riscv64/virt | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.
riscv32/virt | 151 test(s): 122 passed, 7 skipped, 22 failed => status: failure.
s390x/s390-ccw-virtio | 151 test(s): 144 passed, 7 skipped, 0 failed => status: warning.Thanks for all this work. So I've checked a few random archs and trust your tests above to confirm they're correct ;-) I'd just like to get your confirmation regarding statx() support in 4.14 and 4.19, and likely adjust _start_c() according to the last few tests. Also please do prepend to the list the patch that adds the optimize("-Os") to fix _start, and we should be good.
Thanks!
Willy
Re: [PATCH v3 01/11] tools/nolibc: remove the old sys_stat support
In reply to Willy TarreauQuoted 20 lines
Hi Zhangjin,
Quoted 11 lines · Zhangjin Wu
__NR_statxhas been added from v4.10:commit
a528d35e8bfc("statx: Add a system call to make enhanced file info available")It has been supported by all of the platforms since at least from v4.20 and glibc 2.28.
Let's remove the old arch related and dependent
sys_statsupport completely.This is friendly to the future new architecture porting.
As I previously said, I'd like that we at least preserve compatibility with supported stable branches. 4.14 and 4.19 are still supported, so does this mean that if I rebuild my preinit against the updated nolibc it will fail to boot on such older systems for the archs that we support?
Quoted 7 lines
Because if it means that in order to support all currently active kernels, one must only build from an older copy of the lib, that becomes a disservice to its development and usage. Thus if you checked that aarch64, arm, i386, mips, riscv, s390 and
x86_64had already adopted statx by 4.14, then I'm fine and we can dropstat(), but then it must be mentioned in the commit message, because here it's not explicit.
Yeah, we used 'git blame' before and found the last related change is v4.20, but it is not really the first change.
Just read the statx manpage again:
https://man7.org/linux/man-pages/man2/statx.2.html
It shows something about "Linux 4.11, glibc 2.28.
And 'git grep' shows it is true:
$ git grep -r statx v4.11 arch/ include/uapi/asm-generic/unistd.h | grep -E "aarch64|arm|mips|s390|x86|:include/uapi"
v4.11:arch/arm/tools/syscall.tbl:397 common statx sys_statx
v4.11:arch/arm64/include/asm/unistd32.h:#define __NR_statx 397
v4.11:arch/arm64/include/asm/unistd32.h:__SYSCALL(__NR_statx, sys_statx)
v4.11:arch/mips/include/uapi/asm/unistd.h:#define __NR_statx (__NR_Linux + 366)
v4.11:arch/mips/include/uapi/asm/unistd.h:#define __NR_statx (__NR_Linux + 326)
v4.11:arch/mips/include/uapi/asm/unistd.h:#define __NR_statx (__NR_Linux + 330)
v4.11:arch/mips/kernel/scall32-o32.S: PTR sys_statx
v4.11:arch/mips/kernel/scall64-64.S: PTR sys_statx
v4.11:arch/mips/kernel/scall64-n32.S: PTR sys_statx /* 6330 */
v4.11:arch/mips/kernel/scall64-o32.S: PTR sys_statx
v4.11:arch/s390/include/uapi/asm/unistd.h:#define __NR_statx 379
v4.11:arch/s390/kernel/compat_wrapper.c:COMPAT_SYSCALL_WRAP5(statx, int, dfd, const char __user *, path, unsigned, flags, unsigned, mask, struct statx __user *, buffer);
v4.11:arch/s390/kernel/syscalls.S:SYSCALL(sys_statx,compat_sys_statx)
v4.11:arch/x86/entry/syscalls/syscall_32.tbl:383 i386 statx sys_statx
v4.11:arch/x86/entry/syscalls/syscall_64.tbl:332 common statx sys_statx
v4.11:include/uapi/asm-generic/unistd.h:#define __NR_statx 291
v4.11:include/uapi/asm-generic/unistd.h:__SYSCALL(__NR_statx, sys_statx)both riscv and loongarch use the generic unistd.h, so, all of our supported archs should work as-is. riscv itself is added from v4.15, loongarch itself is added from v5.19.
The powerpc we plan to support is from v4.11:
$ git grep -r statx v4.11 arch/powerpc/
v4.11:arch/powerpc/include/asm/systbl.h:SYSCALL(statx)
v4.11:arch/powerpc/include/uapi/asm/unistd.h:#define __NR_statx 383So, let's simply correct the commit message (the old v4.10 has a wrong -1 offset, I wrongly used 'git show a528d35e8bfc:Makefile' before).
sys_statx has been supported from Linux v4.11 and glibc 2.28:
$ git grep -r statx v4.11 arch/ include/uapi/asm-generic/unistd.h
Both riscv (firstly added from v4.15) and loongarch (firstly added from v5.19) use the generic unistd.h, the others are supported from arch/.
Let's remove the old arch related and dependent sys_stat support completely.
Since the current oldest stable branch is v4.14, so, using statx instead of sys_stat preserves compatibility with all of the supported stable branches.
This is friendly to the future new architecture porting.
Best regards,
Zhangjin
Quoted 2 lines
Thanks!
Willy
Willy Tarreau
maintainerRe: [PATCH v3 01/11] tools/nolibc: remove the old sys_stat support
Quoted 27 lines · Zhangjin Wu
Just read the statx manpage again:
https://man7.org/linux/man-pages/man2/statx.2.htmlIt shows something about "Linux 4.11, glibc 2.28.
And 'git grep' shows it is true:
$ git grep -r statx v4.11 arch/ include/uapi/asm-generic/unistd.h | grep -E "aarch64|arm|mips|s390|x86|:include/uapi" v4.11:arch/arm/tools/syscall.tbl:397 common statx sys_statx v4.11:arch/arm64/include/asm/unistd32.h:#define __NR_statx 397 v4.11:arch/arm64/include/asm/unistd32.h:__SYSCALL(__NR_statx, sys_statx) v4.11:arch/mips/include/uapi/asm/unistd.h:#define __NR_statx (__NR_Linux + 366) v4.11:arch/mips/include/uapi/asm/unistd.h:#define __NR_statx (__NR_Linux + 326) v4.11:arch/mips/include/uapi/asm/unistd.h:#define __NR_statx (__NR_Linux + 330) v4.11:arch/mips/kernel/scall32-o32.S: PTR sys_statx v4.11:arch/mips/kernel/scall64-64.S: PTR sys_statx v4.11:arch/mips/kernel/scall64-n32.S: PTR sys_statx /* 6330 */ v4.11:arch/mips/kernel/scall64-o32.S: PTR sys_statx v4.11:arch/s390/include/uapi/asm/unistd.h:#define __NR_statx 379 v4.11:arch/s390/kernel/compat_wrapper.c:COMPAT_SYSCALL_WRAP5(statx, int, dfd, const char __user *, path, unsigned, flags, unsigned, mask, struct statx __user *, buffer); v4.11:arch/s390/kernel/syscalls.S:SYSCALL(sys_statx,compat_sys_statx) v4.11:arch/x86/entry/syscalls/syscall_32.tbl:383 i386 statx sys_statx v4.11:arch/x86/entry/syscalls/syscall_64.tbl:332 common statx sys_statx v4.11:include/uapi/asm-generic/unistd.h:#define __NR_statx 291 v4.11:include/uapi/asm-generic/unistd.h:__SYSCALL(__NR_statx, sys_statx)
Quoted 3 lines
both riscv and loongarch use the generic
unistd.h, so, all of our supported archs should work as-is. riscv itself is added from v4.15, loongarch itself is added from v5.19.
So that's perfect, thank you!
Willy