public inbox for libc-alpha@sourceware.org
 help / color / mirror / Atom feed
From: caiyinyu <caiyinyu@loongson.cn>
To: Adhemerval Zanella Netto <adhemerval.zanella@linaro.org>,
	libc-alpha@sourceware.org
Cc: joseph_myers@mentor.com
Subject: Re: [PATCH v6 08/13] LoongArch: Linux Syscall Interface
Date: Wed, 13 Sep 2023 19:05:25 +0800	[thread overview]
Message-ID: <71e13529-9bb0-f575-d026-6214534c6de1@loongson.cn> (raw)
In-Reply-To: <702f1397-b470-ba44-9d6d-4525a3dbb0b4@linaro.org>

[-- Attachment #1: Type: text/plain, Size: 36251 bytes --]


+
+/* In order to get __set_errno() definition in INLINE_SYSCALL.  */
+#ifndef __ASSEMBLER__
+#include <errno.h>
+#endif
+
+#include <sysdeps/unix/sysdep.h>
+
+#undef SYS_ify
+#define SYS_ify(syscall_name) __NR_##syscall_name
+
+#ifndef __ASSEMBLER__
+
+#define VDSO_NAME "LINUX_2.6"
+#define VDSO_HASH 61765110
+
+/* List of system calls which are supported as vsyscalls.  */
+#define HAVE_CLOCK_GETRES64_VSYSCALL "__vdso_clock_getres"
+#define HAVE_CLOCK_GETTIME64_VSYSCALL "__vdso_clock_gettime"
+#define HAVE_GETTIMEOFDAY_VSYSCALL "__vdso_gettimeofday"

This is used solely if the architecture defines USE_IFUNC_GETTIMEOFDAY,
which requires IFUNC support and an arch-specific gettimeofday.c.

If I understand correctly iFUNC is not really fully supported or has
some issues (due the wrong relocation placement in the plt section).

So I think it wold be better to just remove it on enabled in a future
optimization.

*Removed.*


+#define __SYSCALL_CLOBBERS \
+  "$t0", "$t1", "$t2", "$t3", "$t4", "$t5", "$t6", "$t7", "$t8", "memory"
+
+extern long int __syscall_error (long int neg_errno);
+
+#endif /* ! __ASSEMBLER__ */
+
+/* Pointer mangling is not supported.  */
+#define PTR_MANGLE(var) (void) (var)
+#define PTR_DEMANGLE(var) (void) (var)

Why havent't you enable pointer mangling? It is used to add some hardening on
some internal usage (for instance atexit functions).

*We will add this support in future.*


在 2022/7/14 上午12:19, Adhemerval Zanella Netto 写道:
> Some comments below.
>
> On 08/07/22 03:52, caiyinyu wrote:
>> ---
>>   sysdeps/loongarch/abort-instr.h               |   2 +
>>   sysdeps/loongarch/hp-timing.h                 |  42 +++
>>   .../unix/sysv/linux/loongarch/arch-syscall.h  | 301 ++++++++++++++++
>>   sysdeps/unix/sysv/linux/loongarch/clone.S     |  96 ++++++
>>   sysdeps/unix/sysv/linux/loongarch/clone3.S    |  83 +++++
>>   sysdeps/unix/sysv/linux/loongarch/sysdep.S    |  53 +++
>>   sysdeps/unix/sysv/linux/loongarch/sysdep.h    | 321 ++++++++++++++++++
>>   sysdeps/unix/sysv/linux/loongarch/vfork.S     |  50 +++
>>   8 files changed, 948 insertions(+)
>>   create mode 100644 sysdeps/loongarch/abort-instr.h
>>   create mode 100644 sysdeps/loongarch/hp-timing.h
>>   create mode 100644 sysdeps/unix/sysv/linux/loongarch/arch-syscall.h
>>   create mode 100644 sysdeps/unix/sysv/linux/loongarch/clone.S
>>   create mode 100644 sysdeps/unix/sysv/linux/loongarch/clone3.S
>>   create mode 100644 sysdeps/unix/sysv/linux/loongarch/sysdep.S
>>   create mode 100644 sysdeps/unix/sysv/linux/loongarch/sysdep.h
>>   create mode 100644 sysdeps/unix/sysv/linux/loongarch/vfork.S
>>
>> diff --git a/sysdeps/loongarch/abort-instr.h b/sysdeps/loongarch/abort-instr.h
>> new file mode 100644
>> index 0000000000..46d3ad0871
>> --- /dev/null
>> +++ b/sysdeps/loongarch/abort-instr.h
>> @@ -0,0 +1,2 @@
>> +/* An instruction which should crash any program is a breakpoint.  */
>> +#define ABORT_INSTRUCTION asm ("break 0")
>> diff --git a/sysdeps/loongarch/hp-timing.h b/sysdeps/loongarch/hp-timing.h
>> new file mode 100644
>> index 0000000000..af1da81024
>> --- /dev/null
>> +++ b/sysdeps/loongarch/hp-timing.h
>> @@ -0,0 +1,42 @@
>> +/* High precision, low overhead timing functions.
>> +   Copyright (C) 2022 Free Software Foundation, Inc.
>> +   This file is part of the GNU C Library.
>> +
>> +   The GNU C Library is free software; you can redistribute it and/or
>> +   modify it under the terms of the GNU Lesser General Public
>> +   License as published by the Free Software Foundation; either
>> +   version 2.1 of the License, or (at your option) any later version.
>> +
>> +   The GNU C Library is distributed in the hope that it will be useful,
>> +   but WITHOUT ANY WARRANTY; without even the implied warranty of
>> +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
>> +   Lesser General Public License for more details.
>> +
>> +   You should have received a copy of the GNU Lesser General Public
>> +   License along with the GNU C Library; if not, see
>> +<https://www.gnu.org/licenses/>.  */
>> +
>> +#ifndef _HP_TIMING_H
>> +#define _HP_TIMING_H 1
>> +
>> +/* We always assume having the timestamp register.  */
>> +#define HP_TIMING_AVAIL (1)
>> +#define HP_SMALL_TIMING_AVAIL (1)
>> +
>> +/* We indeed have inlined functions.  */
>> +#define HP_TIMING_INLINE (1)
>> +
>> +/* We use 64bit values for the times.  */
>> +typedef unsigned long long int hp_timing_t;
>> +
>> +/* Read the stable counter.  */
>> +#define HP_TIMING_NOW(Var) \
>> +  ({ \
>> +    unsigned long long int _count; \
>> +    asm volatile ("rdtime.d\t%0,$r0" : "=r" (_count)); \
>> +    (Var) = _count; \
>> +  })
>> +
>> +#include <hp-timing-common.h>
>> +
>> +#endif /* hp-timing.h */
> Ok.
>
>> diff --git a/sysdeps/unix/sysv/linux/loongarch/arch-syscall.h b/sysdeps/unix/sysv/linux/loongarch/arch-syscall.h
>> new file mode 100644
>> index 0000000000..6bb3c8adbc
>> --- /dev/null
>> +++ b/sysdeps/unix/sysv/linux/loongarch/arch-syscall.h
>> @@ -0,0 +1,301 @@
>> +/* AUTOGENERATED by update-syscall-lists.py.  */
>> +#define __NR_accept 202
>> +#define __NR_accept4 242
>> +#define __NR_acct 89
>> +#define __NR_add_key 217
>> +#define __NR_adjtimex 171
>> +#define __NR_bind 200
>> +#define __NR_bpf 280
>> +#define __NR_brk 214
>> +#define __NR_capget 90
>> +#define __NR_capset 91
>> +#define __NR_chdir 49
>> +#define __NR_chroot 51
>> +#define __NR_clock_adjtime 266
>> +#define __NR_clock_getres 114
>> +#define __NR_clock_gettime 113
>> +#define __NR_clock_nanosleep 115
>> +#define __NR_clock_settime 112
>> +#define __NR_clone 220
>> +#define __NR_clone3 435
>> +#define __NR_close 57
>> +#define __NR_close_range 436
>> +#define __NR_connect 203
>> +#define __NR_copy_file_range 285
>> +#define __NR_delete_module 106
>> +#define __NR_dup 23
>> +#define __NR_dup3 24
>> +#define __NR_epoll_create1 20
>> +#define __NR_epoll_ctl 21
>> +#define __NR_epoll_pwait 22
>> +#define __NR_epoll_pwait2 441
>> +#define __NR_eventfd2 19
>> +#define __NR_execve 221
>> +#define __NR_execveat 281
>> +#define __NR_exit 93
>> +#define __NR_exit_group 94
>> +#define __NR_faccessat 48
>> +#define __NR_faccessat2 439
>> +#define __NR_fadvise64 223
>> +#define __NR_fallocate 47
>> +#define __NR_fanotify_init 262
>> +#define __NR_fanotify_mark 263
>> +#define __NR_fchdir 50
>> +#define __NR_fchmod 52
>> +#define __NR_fchmodat 53
>> +#define __NR_fchown 55
>> +#define __NR_fchownat 54
>> +#define __NR_fcntl 25
>> +#define __NR_fdatasync 83
>> +#define __NR_fgetxattr 10
>> +#define __NR_finit_module 273
>> +#define __NR_flistxattr 13
>> +#define __NR_flock 32
>> +#define __NR_fremovexattr 16
>> +#define __NR_fsconfig 431
>> +#define __NR_fsetxattr 7
>> +#define __NR_fsmount 432
>> +#define __NR_fsopen 430
>> +#define __NR_fspick 433
>> +#define __NR_fstatfs 44
>> +#define __NR_fsync 82
>> +#define __NR_ftruncate 46
>> +#define __NR_futex 98
>> +#define __NR_futex_waitv 449
>> +#define __NR_get_mempolicy 236
>> +#define __NR_get_robust_list 100
>> +#define __NR_getcpu 168
>> +#define __NR_getcwd 17
>> +#define __NR_getdents64 61
>> +#define __NR_getegid 177
>> +#define __NR_geteuid 175
>> +#define __NR_getgid 176
>> +#define __NR_getgroups 158
>> +#define __NR_getitimer 102
>> +#define __NR_getpeername 205
>> +#define __NR_getpgid 155
>> +#define __NR_getpid 172
>> +#define __NR_getppid 173
>> +#define __NR_getpriority 141
>> +#define __NR_getrandom 278
>> +#define __NR_getresgid 150
>> +#define __NR_getresuid 148
>> +#define __NR_getrusage 165
>> +#define __NR_getsid 156
>> +#define __NR_getsockname 204
>> +#define __NR_getsockopt 209
>> +#define __NR_gettid 178
>> +#define __NR_gettimeofday 169
>> +#define __NR_getuid 174
>> +#define __NR_getxattr 8
>> +#define __NR_init_module 105
>> +#define __NR_inotify_add_watch 27
>> +#define __NR_inotify_init1 26
>> +#define __NR_inotify_rm_watch 28
>> +#define __NR_io_cancel 3
>> +#define __NR_io_destroy 1
>> +#define __NR_io_getevents 4
>> +#define __NR_io_pgetevents 292
>> +#define __NR_io_setup 0
>> +#define __NR_io_submit 2
>> +#define __NR_io_uring_enter 426
>> +#define __NR_io_uring_register 427
>> +#define __NR_io_uring_setup 425
>> +#define __NR_ioctl 29
>> +#define __NR_ioprio_get 31
>> +#define __NR_ioprio_set 30
>> +#define __NR_kcmp 272
>> +#define __NR_kexec_file_load 294
>> +#define __NR_kexec_load 104
>> +#define __NR_keyctl 219
>> +#define __NR_kill 129
>> +#define __NR_landlock_add_rule 445
>> +#define __NR_landlock_create_ruleset 444
>> +#define __NR_landlock_restrict_self 446
>> +#define __NR_lgetxattr 9
>> +#define __NR_linkat 37
>> +#define __NR_listen 201
>> +#define __NR_listxattr 11
>> +#define __NR_llistxattr 12
>> +#define __NR_lookup_dcookie 18
>> +#define __NR_lremovexattr 15
>> +#define __NR_lseek 62
>> +#define __NR_lsetxattr 6
>> +#define __NR_madvise 233
>> +#define __NR_mbind 235
>> +#define __NR_membarrier 283
>> +#define __NR_memfd_create 279
>> +#define __NR_migrate_pages 238
>> +#define __NR_mincore 232
>> +#define __NR_mkdirat 34
>> +#define __NR_mknodat 33
>> +#define __NR_mlock 228
>> +#define __NR_mlock2 284
>> +#define __NR_mlockall 230
>> +#define __NR_mmap 222
>> +#define __NR_mount 40
>> +#define __NR_mount_setattr 442
>> +#define __NR_move_mount 429
>> +#define __NR_move_pages 239
>> +#define __NR_mprotect 226
>> +#define __NR_mq_getsetattr 185
>> +#define __NR_mq_notify 184
>> +#define __NR_mq_open 180
>> +#define __NR_mq_timedreceive 183
>> +#define __NR_mq_timedsend 182
>> +#define __NR_mq_unlink 181
>> +#define __NR_mremap 216
>> +#define __NR_msgctl 187
>> +#define __NR_msgget 186
>> +#define __NR_msgrcv 188
>> +#define __NR_msgsnd 189
>> +#define __NR_msync 227
>> +#define __NR_munlock 229
>> +#define __NR_munlockall 231
>> +#define __NR_munmap 215
>> +#define __NR_name_to_handle_at 264
>> +#define __NR_nanosleep 101
>> +#define __NR_nfsservctl 42
>> +#define __NR_open_by_handle_at 265
>> +#define __NR_open_tree 428
>> +#define __NR_openat 56
>> +#define __NR_openat2 437
>> +#define __NR_perf_event_open 241
>> +#define __NR_personality 92
>> +#define __NR_pidfd_getfd 438
>> +#define __NR_pidfd_open 434
>> +#define __NR_pidfd_send_signal 424
>> +#define __NR_pipe2 59
>> +#define __NR_pivot_root 41
>> +#define __NR_pkey_alloc 289
>> +#define __NR_pkey_free 290
>> +#define __NR_pkey_mprotect 288
>> +#define __NR_ppoll 73
>> +#define __NR_prctl 167
>> +#define __NR_pread64 67
>> +#define __NR_preadv 69
>> +#define __NR_preadv2 286
>> +#define __NR_prlimit64 261
>> +#define __NR_process_madvise 440
>> +#define __NR_process_mrelease 448
>> +#define __NR_process_vm_readv 270
>> +#define __NR_process_vm_writev 271
>> +#define __NR_pselect6 72
>> +#define __NR_ptrace 117
>> +#define __NR_pwrite64 68
>> +#define __NR_pwritev 70
>> +#define __NR_pwritev2 287
>> +#define __NR_quotactl 60
>> +#define __NR_quotactl_fd 443
>> +#define __NR_read 63
>> +#define __NR_readahead 213
>> +#define __NR_readlinkat 78
>> +#define __NR_readv 65
>> +#define __NR_reboot 142
>> +#define __NR_recvfrom 207
>> +#define __NR_recvmmsg 243
>> +#define __NR_recvmsg 212
>> +#define __NR_remap_file_pages 234
>> +#define __NR_removexattr 14
>> +#define __NR_renameat2 276
>> +#define __NR_request_key 218
>> +#define __NR_restart_syscall 128
>> +#define __NR_rseq 293
>> +#define __NR_rt_sigaction 134
>> +#define __NR_rt_sigpending 136
>> +#define __NR_rt_sigprocmask 135
>> +#define __NR_rt_sigqueueinfo 138
>> +#define __NR_rt_sigreturn 139
>> +#define __NR_rt_sigsuspend 133
>> +#define __NR_rt_sigtimedwait 137
>> +#define __NR_rt_tgsigqueueinfo 240
>> +#define __NR_sched_get_priority_max 125
>> +#define __NR_sched_get_priority_min 126
>> +#define __NR_sched_getaffinity 123
>> +#define __NR_sched_getattr 275
>> +#define __NR_sched_getparam 121
>> +#define __NR_sched_getscheduler 120
>> +#define __NR_sched_rr_get_interval 127
>> +#define __NR_sched_setaffinity 122
>> +#define __NR_sched_setattr 274
>> +#define __NR_sched_setparam 118
>> +#define __NR_sched_setscheduler 119
>> +#define __NR_sched_yield 124
>> +#define __NR_seccomp 277
>> +#define __NR_semctl 191
>> +#define __NR_semget 190
>> +#define __NR_semop 193
>> +#define __NR_semtimedop 192
>> +#define __NR_sendfile 71
>> +#define __NR_sendmmsg 269
>> +#define __NR_sendmsg 211
>> +#define __NR_sendto 206
>> +#define __NR_set_mempolicy 237
>> +#define __NR_set_mempolicy_home_node 450
>> +#define __NR_set_robust_list 99
>> +#define __NR_set_tid_address 96
>> +#define __NR_setdomainname 162
>> +#define __NR_setfsgid 152
>> +#define __NR_setfsuid 151
>> +#define __NR_setgid 144
>> +#define __NR_setgroups 159
>> +#define __NR_sethostname 161
>> +#define __NR_setitimer 103
>> +#define __NR_setns 268
>> +#define __NR_setpgid 154
>> +#define __NR_setpriority 140
>> +#define __NR_setregid 143
>> +#define __NR_setresgid 149
>> +#define __NR_setresuid 147
>> +#define __NR_setreuid 145
>> +#define __NR_setsid 157
>> +#define __NR_setsockopt 208
>> +#define __NR_settimeofday 170
>> +#define __NR_setuid 146
>> +#define __NR_setxattr 5
>> +#define __NR_shmat 196
>> +#define __NR_shmctl 195
>> +#define __NR_shmdt 197
>> +#define __NR_shmget 194
>> +#define __NR_shutdown 210
>> +#define __NR_sigaltstack 132
>> +#define __NR_signalfd4 74
>> +#define __NR_socket 198
>> +#define __NR_socketpair 199
>> +#define __NR_splice 76
>> +#define __NR_statfs 43
>> +#define __NR_statx 291
>> +#define __NR_swapoff 225
>> +#define __NR_swapon 224
>> +#define __NR_symlinkat 36
>> +#define __NR_sync 81
>> +#define __NR_sync_file_range 84
>> +#define __NR_syncfs 267
>> +#define __NR_sysinfo 179
>> +#define __NR_syslog 116
>> +#define __NR_tee 77
>> +#define __NR_tgkill 131
>> +#define __NR_timer_create 107
>> +#define __NR_timer_delete 111
>> +#define __NR_timer_getoverrun 109
>> +#define __NR_timer_gettime 108
>> +#define __NR_timer_settime 110
>> +#define __NR_timerfd_create 85
>> +#define __NR_timerfd_gettime 87
>> +#define __NR_timerfd_settime 86
>> +#define __NR_times 153
>> +#define __NR_tkill 130
>> +#define __NR_truncate 45
>> +#define __NR_umask 166
>> +#define __NR_umount2 39
>> +#define __NR_uname 160
>> +#define __NR_unlinkat 35
>> +#define __NR_unshare 97
>> +#define __NR_userfaultfd 282
>> +#define __NR_utimensat 88
>> +#define __NR_vhangup 58
>> +#define __NR_vmsplice 75
>> +#define __NR_wait4 260
>> +#define __NR_waitid 95
>> +#define __NR_write 64
>> +#define __NR_writev 66
> Ok, it matches the output of 'make update-syscall-lists' command.
>
>> diff --git a/sysdeps/unix/sysv/linux/loongarch/clone.S b/sysdeps/unix/sysv/linux/loongarch/clone.S
>> new file mode 100644
>> index 0000000000..c73245be82
>> --- /dev/null
>> +++ b/sysdeps/unix/sysv/linux/loongarch/clone.S
>> @@ -0,0 +1,96 @@
>> +/* The clone syscall wrapper.
>> +   Copyright (C) 2022 Free Software Foundation, Inc.
>> +
>> +   This file is part of the GNU C Library.
>> +
>> +   The GNU C Library is free software; you can redistribute it and/or
>> +   modify it under the terms of the GNU Lesser General Public
>> +   License as published by the Free Software Foundation; either
>> +   version 2.1 of the License, or (at your option) any later version.
>> +
>> +   The GNU C Library is distributed in the hope that it will be useful,
>> +   but WITHOUT ANY WARRANTY; without even the implied warranty of
>> +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
>> +   Lesser General Public License for more details.
>> +
>> +   You should have received a copy of the GNU Lesser General Public
>> +   License along with the GNU C Library.  If not, see
>> +<https://www.gnu.org/licenses/>.  */
>> +
>> +/* clone() is even more special than fork() as it mucks with stacks
>> +   and invokes a function in the right context after its all over.  */
>> +
>> +#include <sys/asm.h>
>> +#include <sysdep.h>
>> +#define _ERRNO_H  1
>> +#include <bits/errno.h>
>> +#include <tls.h>
>> +#include "tcb-offsets.h"
>> +
>> +/* int clone(int (*fn)(void *arg), void *child_stack, int flags, void *arg,
>> +   void *parent_tidptr, void *tls, void *child_tidptr) */
>> +
>> +ENTRY (__clone)
>> +
>> +	/* Align stack to 16.  */
>> +	BSTRINS		a1, zero, 3, 0
>> +
>> +	/* Sanity check arguments.  */
>> +	beqz		a0, L (invalid) /* No NULL function pointers.  */
>> +	beqz		a1, L (invalid) /* No NULL stack pointers.  */
>> +
>> +	ADDI 		a1, a1, -16 /* Reserve argument save space.  */
>> +	REG_S		a0, a1, 0   /* Save function pointer.  */
>> +	REG_S		a3, a1, SZREG   /* Save argument pointer.  */
>> +
>> +	/* The syscall expects the args to be in different slots.  */
>> +	or		a0, a2, zero
>> +	or		a2, a4, zero
>> +	or		a3, a6, zero
>> +	or		a4, a5, zero
>> +
>> +	/* Do the system call.  */
>> +	LI		a7,__NR_clone
>> +	syscall		0
>> +
>> +	blt		a0, zero ,L (error)
>> +	beqz		a0,L (thread_start)
>> +
>> +	/* Successful return from the parent.  */
>> +	ret
>> +
>> +L (invalid):
>> +	LI		a0, -EINVAL
>> +
>> +	/* Something bad happened -- no child created.  */
>> +L (error):
>> +	b		__syscall_error
>> +
>> +END (__clone)
>> +
>> +/* Load up the arguments to the function.  Put this block of code in
>> +   its own function so that we can terminate the stack trace with our
>> +   debug info.  */
>> +ENTRY (__thread_start)
>> +L (thread_start):
>> +
>> +/* Terminate call stack by noting ra is undefined.  Use a dummy
>> +   .cfi_label to force starting the FDE.  */
>> +	.cfi_label .Ldummy
>> +	cfi_undefined (1)
>> +
>> +	/* Restore the arg for user's function.  */
>> +	REG_L		a1, sp, 0   /* Function pointer.  */
>> +	REG_L		a0, sp, SZREG   /* Argument pointer.  */
>> +
>> +	/* Call the user's function.  */
>> +	jirl		ra, a1, 0
>> +
>> +	/* Call exit with the function's return value.  */
>> +	LI		a7, __NR_exit
>> +	syscall		0
>> +
>> +	END (__thread_start)
>> +
>> +libc_hidden_def (__clone)
>> +weak_alias (__clone, clone)
> Ok.
>
>> diff --git a/sysdeps/unix/sysv/linux/loongarch/clone3.S b/sysdeps/unix/sysv/linux/loongarch/clone3.S
>> new file mode 100644
>> index 0000000000..38be4c809b
>> --- /dev/null
>> +++ b/sysdeps/unix/sysv/linux/loongarch/clone3.S
>> @@ -0,0 +1,83 @@
>> +/* The clone3 syscall wrapper.
>> +   Copyright (C) 2022 Free Software Foundation, Inc.
>> +
>> +   This file is part of the GNU C Library.
>> +
>> +   The GNU C Library is free software; you can redistribute it and/or
>> +   modify it under the terms of the GNU Lesser General Public
>> +   License as published by the Free Software Foundation; either
>> +   version 2.1 of the License, or (at your option) any later version.
>> +
>> +   The GNU C Library is distributed in the hope that it will be useful,
>> +   but WITHOUT ANY WARRANTY; without even the implied warranty of
>> +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
>> +   Lesser General Public License for more details.
>> +
>> +   You should have received a copy of the GNU Lesser General Public
>> +   License along with the GNU C Library.  If not, see
>> +<https://www.gnu.org/licenses/>.  */
>> +
>> +/* clone3() is even more special than fork() as it mucks with stacks
>> +   and invokes a function in the right context after its all over.  */
>> +
>> +#include <sys/asm.h>
>> +#include <sysdep.h>
>> +#define _ERRNO_H  1
>> +#include <bits/errno.h>
>> +#include <tls.h>
>> +#include "tcb-offsets.h"
>> +
>> +/* int clone3(struct clone_args *cl_args, size_t size,
>> +   int (*func)(void *arg), void *arg); */
>> +
>> +ENTRY (__clone3)
>> +
>> +	/* Sanity check arguments.  */
>> +	beqz		a0, L (invalid)	/* No NULL cl_args pointer.  */
>> +	beqz		a2, L (invalid)	/* No NULL function pointer.  */
>> +
>> +	/* Do the system call.  */
>> +	LI		a7, __NR_clone3
>> +	syscall		0
>> +
>> +	blt		a0, zero ,L (error)
>> +	beqz		a0, L (thread_start3)
>> +
>> +	/* Successful return from the parent.  */
>> +	ret
>> +
>> +L (invalid):
>> +	LI		a0, -EINVAL
>> +
>> +	/* Something bad happened -- no child created.  */
>> +L (error):
>> +	b		__syscall_error
>> +
>> +END (__clone3)
>> +
>> +/* Load up the arguments to the function.  Put this block of code in
>> +   its own function so that we can terminate the stack trace with our
>> +   debug info.  */
>> +ENTRY (__thread_start3)
>> +L (thread_start3):
>> +
>> +/* Terminate call stack by noting ra is undefined.  Use a dummy
>> +   .cfi_label to force starting the FDE.  */
>> +	.cfi_label .Ldummy
>> +	cfi_undefined (1)
>> +
>> +	/* Align stack to 16.  */
>> +	BSTRINS		sp, zero, 3, 0
>> +
>> +	/* Set up arguments for the function call.  */
>> +	move		a0, a3		/* Argument.  */
>> +	jirl		ra, a2, 0	/* Call function.  */
>> +
>> +	/* Call exit with the function's return value.  */
>> +	LI		a7, __NR_exit
>> +	syscall		0
>> +
>> +	END (__thread_start3)
>> +
>> +libc_hidden_def (__clone3)
>> +weak_alias (__clone3, clone3)
> Ok.
>
>> diff --git a/sysdeps/unix/sysv/linux/loongarch/sysdep.S b/sysdeps/unix/sysv/linux/loongarch/sysdep.S
>> new file mode 100644
>> index 0000000000..a4db2a97b5
>> --- /dev/null
>> +++ b/sysdeps/unix/sysv/linux/loongarch/sysdep.S
>> @@ -0,0 +1,53 @@
>> +/* syscall error handlers
>> +   Copyright (C) 2022 Free Software Foundation, Inc.
>> +   This file is part of the GNU C Library.
>> +
>> +   The GNU C Library is free software; you can redistribute it and/or
>> +   modify it under the terms of the GNU Lesser General Public
>> +   License as published by the Free Software Foundation; either
>> +   version 2.1 of the License, or (at your option) any later version.
>> +
>> +   The GNU C Library is distributed in the hope that it will be useful,
>> +   but WITHOUT ANY WARRANTY; without even the implied warranty of
>> +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
>> +   Lesser General Public License for more details.
>> +
>> +   You should have received a copy of the GNU Lesser General Public
>> +   License along with the GNU C Library.  If not, see
>> +<https://www.gnu.org/licenses/>.  */
>> +
>> +#include <sysdep.h>
>> +
>> +#if IS_IN (libc)
>> +# define errno __libc_errno
>> +#endif
>> +
>> +ENTRY (__syscall_error)
>> +/* Fall through to __syscall_set_errno */
>> +END (__syscall_error)
>> +
>> +/* Non-standard calling convention: argument in a0, return address in t0,
>> +   and clobber only t1.
>> + */
>> +ENTRY (__syscall_set_errno)
>> +
>> +/* We got here because a0 < 0, but only codes in the range [-4095, -1]
>> +   represent errors.  Otherwise, just return the result normally.
>> + */
>> +	li.d		t1, -4096
>> +	bgeu		t1, a0, L (out)
>> +	sub.w		a0, zero, a0
>> +#if RTLD_PRIVATE_ERRNO
>> +	la		t1, rtld_errno
>> +#elif defined(__PIC__)
>> +	la.tls.ie	t1, errno
>> +	add.d		t1, tp, t1
>> +#else
>> +	la.tls.le	t1, errno
>> +	add.d		t1, tp, t1
>> +#endif
>> +	st.w		a0, t1, 0
>> +	li.d		a0, -1
>> +L (out):
>> +	ret
>> +END (__syscall_set_errno)
> Ok.
>
>> diff --git a/sysdeps/unix/sysv/linux/loongarch/sysdep.h b/sysdeps/unix/sysv/linux/loongarch/sysdep.h
>> new file mode 100644
>> index 0000000000..a90d2b65ed
>> --- /dev/null
>> +++ b/sysdeps/unix/sysv/linux/loongarch/sysdep.h
>> @@ -0,0 +1,321 @@
>> +/* Assembly macros for LoongArch.
>> +   Copyright (C) 2022 Free Software Foundation, Inc.
>> +   This file is part of the GNU C Library.
>> +
>> +   The GNU C Library is free software; you can redistribute it and/or
>> +   modify it under the terms of the GNU Lesser General Public
>> +   License as published by the Free Software Foundation; either
>> +   version 2.1 of the License, or (at your option) any later version.
>> +
>> +   The GNU C Library is distributed in the hope that it will be useful,
>> +   but WITHOUT ANY WARRANTY; without even the implied warranty of
>> +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
>> +   Lesser General Public License for more details.
>> +
>> +   You should have received a copy of the GNU Lesser General Public
>> +   License along with the GNU C Library.  If not, see
>> +<https://www.gnu.org/licenses/>.  */
>> +
>> +#ifndef _LINUX_LOONGARCH_SYSDEP_H
>> +#define _LINUX_LOONGARCH_SYSDEP_H 1
>> +
>> +#include <sysdeps/unix/sysv/linux/generic/sysdep.h>
>> +#include <tls.h>
>> +
>> +#ifdef __ASSEMBLER__
>> +
>> +#include <sys/asm.h>
>> +#define ret jirl zero, ra, 0
>> +#define L(label) .L##label
>> +
>> +/* Performs a system call, handling errors by setting errno.  Linux indicates
>> +   errors by setting a0 to a value between -1 and -4095.  */
>> +#undef PSEUDO
>> +#define PSEUDO(name, syscall_name, args) \
>> +  ENTRY (name); \
>> +  li.d a7, SYS_ify (syscall_name); \
>> +  syscall 0; \
>> +  li.d a7, -4096; \
>> +  bltu a7, a0, .Lsyscall_error##name;
>> +
>> +#undef PSEUDO_END
>> +#define PSEUDO_END(sym) \
>> +  SYSCALL_ERROR_HANDLER (sym); \
>> +  ret; \
>> +  END (sym);
>> +
>> +#if !IS_IN(libc)
>> +#if RTLD_PRIVATE_ERRNO
>> +
>> +#define SYSCALL_ERROR_HANDLER(name) \
>> +  .Lsyscall_error##name : la t0, rtld_errno; \
>> +  sub.w a0, zero, a0; \
>> +  st.w a0, t0, 0; \
>> +  li.d a0, -1;
>> +
>> +#else
>> +
>> +#define SYSCALL_ERROR_HANDLER(name) \
>> +  .Lsyscall_error##name : la.tls.ie t0, errno; \
>> +  add.d t0, tp, t0; \
>> +  sub.w a0, zero, a0; \
>> +  st.w a0, t0, 0; \
>> +  li.d a0, -1;
>> +
>> +#endif
>> +#else
>> +
>> +#define SYSCALL_ERROR_HANDLER(name) .Lsyscall_error##name : b __syscall_error;
>> +
>> +#endif
>> +
>> +/* Performs a system call, not setting errno.  */
>> +#undef PSEUDO_NEORRNO
>> +#define PSEUDO_NOERRNO(name, syscall_name, args) \
>> +  ENTRY (name); \
>> +  li.d a7, SYS_ify (syscall_name); \
>> +  syscall 0;
>> +
>> +#undef PSEUDO_END_NOERRNO
>> +#define PSEUDO_END_NOERRNO(name) END (name);
>> +
>> +#undef ret_NOERRNO
>> +#define ret_NOERRNO ret
>> +
>> +/* Perfroms a system call, returning the error code.  */
>> +#undef PSEUDO_ERRVAL
>> +#define PSEUDO_ERRVAL(name, syscall_name, args) \
>> +  PSEUDO_NOERRNO (name, syscall_name, args); \
>> +  slli.d a0, a0, 32; \
>> +  srai.d a0, a0, 32; /* sign_ext */ \
>> +  sub.d a0, zero, a0;
>> +
>> +#undef PSEUDO_END_ERRVAL
>> +#define PSEUDO_END_ERRVAL(name) END (name);
>> +
>> +#undef ret_ERRVAL
>> +#define ret_ERRVAL ret
>> +
>> +#endif /* __ASSEMBLER__ */
>> +
>> +/* In order to get __set_errno() definition in INLINE_SYSCALL.  */
>> +#ifndef __ASSEMBLER__
>> +#include <errno.h>
>> +#endif
>> +
>> +#include <sysdeps/unix/sysdep.h>
>> +
>> +#undef SYS_ify
>> +#define SYS_ify(syscall_name) __NR_##syscall_name
>> +
>> +#ifndef __ASSEMBLER__
>> +
>> +#define VDSO_NAME "LINUX_2.6"
>> +#define VDSO_HASH 61765110
>> +
>> +/* List of system calls which are supported as vsyscalls.  */
>> +#define HAVE_CLOCK_GETRES64_VSYSCALL "__vdso_clock_getres"
>> +#define HAVE_CLOCK_GETTIME64_VSYSCALL "__vdso_clock_gettime"
>> +#define HAVE_GETTIMEOFDAY_VSYSCALL "__vdso_gettimeofday"
> This is used solely if the architecture defines USE_IFUNC_GETTIMEOFDAY,
> which requires IFUNC support and an arch-specific gettimeofday.c.
>
> If I understand correctly iFUNC is not really fully supported or has
> some issues (due the wrong relocation placement in the plt section).
>
> So I think it wold be better to just remove it on enabled in a future
> optimization.
>
>> +#define HAVE_GETCPU_VSYSCALL "__vdso_getcpu"
>> +
>> +#define HAVE_CLONE3_WRAPPER 1
>> +
>> +#define INTERNAL_SYSCALL(name, nr, args...) \
>> +  internal_syscall##nr (SYS_ify (name), args)
>> +
>> +#define INTERNAL_SYSCALL_NCS(number, nr, args...) \
>> +  internal_syscall##nr (number, args)
>> +
>> +#define internal_syscall0(number, dummy...) \
>> +  ({ \
>> +    long int _sys_result; \
>> +\
>> +    { \
>> +      register long int __a7 asm("$a7") = number; \
> Missing space after asm, there are more occurances on other macros in
> this file.
>
>> +      register long int __a0 asm("$a0"); \
>> +      __asm__ volatile("syscall	0\n\t" \
>> +		       : "=r"(__a0) \
>> +		       : "r"(__a7) \
>> +		       : __SYSCALL_CLOBBERS); \
>> +      _sys_result = __a0; \
>> +    } \
>> +    _sys_result; \
>> +  })
>> +
>> +#define internal_syscall1(number, arg0) \
>> +  ({ \
>> +    long int _sys_result; \
>> +\
>> +    { \
>> +      long int _arg0 = (long int) (arg0); \
>> +      register long int __a7 asm("$a7") = number; \
>> +      register long int __a0 asm("$a0") = _arg0; \
>> +      __asm__ volatile("syscall	0\n\t" \
>> +		       : "+r"(__a0) \
>> +		       : "r"(__a7) \
>> +		       : __SYSCALL_CLOBBERS); \
>> +      _sys_result = __a0; \
>> +    } \
>> +    _sys_result; \
>> +  })
>> +
>> +#define internal_syscall2(number, arg0, arg1) \
>> +  ({ \
>> +    long int _sys_result; \
>> +\
>> +    { \
>> +      long int _arg0 = (long int) (arg0); \
>> +      long int _arg1 = (long int) (arg1); \
>> +      register long int __a7 asm("$a7") = number; \
>> +      register long int __a0 asm("$a0") = _arg0; \
>> +      register long int __a1 asm("$a1") = _arg1; \
>> +      __asm__ volatile("syscall	0\n\t" \
>> +		       : "+r"(__a0) \
>> +		       : "r"(__a7), "r"(__a1) \
>> +		       : __SYSCALL_CLOBBERS); \
>> +      _sys_result = __a0; \
>> +    } \
>> +    _sys_result; \
>> +  })
>> +
>> +#define internal_syscall3(number, arg0, arg1, arg2) \
>> +  ({ \
>> +    long int _sys_result; \
>> +\
>> +    { \
>> +      long int _arg0 = (long int) (arg0); \
>> +      long int _arg1 = (long int) (arg1); \
>> +      long int _arg2 = (long int) (arg2); \
>> +      register long int __a7 asm("$a7") = number; \
>> +      register long int __a0 asm("$a0") = _arg0; \
>> +      register long int __a1 asm("$a1") = _arg1; \
>> +      register long int __a2 asm("$a2") = _arg2; \
>> +      __asm__ volatile("syscall	0\n\t" \
>> +		       : "+r"(__a0) \
>> +		       : "r"(__a7), "r"(__a1), "r"(__a2) \
>> +		       : __SYSCALL_CLOBBERS); \
>> +      _sys_result = __a0; \
>> +    } \
>> +    _sys_result; \
>> +  })
>> +
>> +#define internal_syscall4(number, arg0, arg1, arg2, arg3) \
>> +  ({ \
>> +    long int _sys_result; \
>> +\
>> +    { \
>> +      long int _arg0 = (long int) (arg0); \
>> +      long int _arg1 = (long int) (arg1); \
>> +      long int _arg2 = (long int) (arg2); \
>> +      long int _arg3 = (long int) (arg3); \
>> +      register long int __a7 asm("$a7") = number; \
>> +      register long int __a0 asm("$a0") = _arg0; \
>> +      register long int __a1 asm("$a1") = _arg1; \
>> +      register long int __a2 asm("$a2") = _arg2; \
>> +      register long int __a3 asm("$a3") = _arg3; \
>> +      __asm__ volatile("syscall	0\n\t" \
>> +		       : "+r"(__a0) \
>> +		       : "r"(__a7), "r"(__a1), "r"(__a2), "r"(__a3) \
>> +		       : __SYSCALL_CLOBBERS); \
>> +      _sys_result = __a0; \
>> +    } \
>> +    _sys_result; \
>> +  })
>> +
>> +#define internal_syscall5(number, arg0, arg1, arg2, arg3, arg4) \
>> +  ({ \
>> +    long int _sys_result; \
>> +\
>> +    { \
>> +      long int _arg0 = (long int) (arg0); \
>> +      long int _arg1 = (long int) (arg1); \
>> +      long int _arg2 = (long int) (arg2); \
>> +      long int _arg3 = (long int) (arg3); \
>> +      long int _arg4 = (long int) (arg4); \
>> +      register long int __a7 asm("$a7") = number; \
>> +      register long int __a0 asm("$a0") = _arg0; \
>> +      register long int __a1 asm("$a1") = _arg1; \
>> +      register long int __a2 asm("$a2") = _arg2; \
>> +      register long int __a3 asm("$a3") = _arg3; \
>> +      register long int __a4 asm("$a4") = _arg4; \
>> +      __asm__ volatile("syscall	0\n\t" \
>> +		       : "+r"(__a0) \
>> +		       : "r"(__a7), "r"(__a1), "r"(__a2), \
>> +			 "r"(__a3), "r"(__a4) \
>> +		       : __SYSCALL_CLOBBERS); \
>> +      _sys_result = __a0; \
>> +    } \
>> +    _sys_result; \
>> +  })
>> +
>> +#define internal_syscall6(number, arg0, arg1, arg2, arg3, arg4, arg5) \
>> +  ({ \
>> +    long int _sys_result; \
>> +\
>> +    { \
>> +      long int _arg0 = (long int) (arg0); \
>> +      long int _arg1 = (long int) (arg1); \
>> +      long int _arg2 = (long int) (arg2); \
>> +      long int _arg3 = (long int) (arg3); \
>> +      long int _arg4 = (long int) (arg4); \
>> +      long int _arg5 = (long int) (arg5); \
>> +      register long int __a7 asm("$a7") = number; \
>> +      register long int __a0 asm("$a0") = _arg0; \
>> +      register long int __a1 asm("$a1") = _arg1; \
>> +      register long int __a2 asm("$a2") = _arg2; \
>> +      register long int __a3 asm("$a3") = _arg3; \
>> +      register long int __a4 asm("$a4") = _arg4; \
>> +      register long int __a5 asm("$a5") = _arg5; \
>> +      __asm__ volatile("syscall	0\n\t" \
>> +		       : "+r"(__a0) \
>> +		       : "r"(__a7), "r"(__a1), "r"(__a2), "r"(__a3), \
>> +			 "r"(__a4), "r"(__a5) \
>> +		       : __SYSCALL_CLOBBERS); \
>> +      _sys_result = __a0; \
>> +    } \
>> +    _sys_result; \
>> +  })
>> +
>> +#define internal_syscall7(number, arg0, arg1, arg2, arg3, arg4, arg5, arg6) \
>> +  ({ \
>> +    long int _sys_result; \
>> +\
>> +    { \
>> +      long int _arg0 = (long int) (arg0); \
>> +      long int _arg1 = (long int) (arg1); \
>> +      long int _arg2 = (long int) (arg2); \
>> +      long int _arg3 = (long int) (arg3); \
>> +      long int _arg4 = (long int) (arg4); \
>> +      long int _arg5 = (long int) (arg5); \
>> +      long int _arg6 = (long int) (arg6); \
>> +      register long int __a7 asm("$a7") = number; \
>> +      register long int __a0 asm("$a0") = _arg0; \
>> +      register long int __a1 asm("$a1") = _arg1; \
>> +      register long int __a2 asm("$a2") = _arg2; \
>> +      register long int __a3 asm("$a3") = _arg3; \
>> +      register long int __a4 asm("$a4") = _arg4; \
>> +      register long int __a5 asm("$a5") = _arg5; \
>> +      register long int __a6 asm("$a6") = _arg6; \
>> +      __asm__ volatile("syscall	0\n\t" \
>> +		       : "+r"(__a0) \
>> +		       : "r"(__a7), "r"(__a1), "r"(__a2), "r"(__a3), \
>> +			 "r"(__a4), "r"(__a5), "r"(__a6) \
>> +		       : __SYSCALL_CLOBBERS); \
>> +      _sys_result = __a0; \
>> +    } \
>> +    _sys_result; \
>> +  })
>> +
>> +#define __SYSCALL_CLOBBERS \
>> +  "$t0", "$t1", "$t2", "$t3", "$t4", "$t5", "$t6", "$t7", "$t8", "memory"
>> +
>> +extern long int __syscall_error (long int neg_errno);
>> +
>> +#endif /* ! __ASSEMBLER__ */
>> +
>> +/* Pointer mangling is not supported.  */
>> +#define PTR_MANGLE(var) (void) (var)
>> +#define PTR_DEMANGLE(var) (void) (var)
> Why havent't you enable pointer mangling? It is used to add some hardening on
> some internal usage (for instance atexit functions).
>
>> +
>> +#endif /* linux/loongarch/sysdep.h */
>> diff --git a/sysdeps/unix/sysv/linux/loongarch/vfork.S b/sysdeps/unix/sysv/linux/loongarch/vfork.S
>> new file mode 100644
>> index 0000000000..fcc84e3549
>> --- /dev/null
>> +++ b/sysdeps/unix/sysv/linux/loongarch/vfork.S
>> @@ -0,0 +1,50 @@
>> +/* vfork for Linux.
>> +   Copyright (C) 2022 Free Software Foundation, Inc.
>> +
>> +   This file is part of the GNU C Library.
>> +
>> +   The GNU C Library is free software; you can redistribute it and/or
>> +   modify it under the terms of the GNU Lesser General Public License as
>> +   published by the Free Software Foundation; either version 2.1 of the
>> +   License, or (at your option) any later version.
>> +
>> +   The GNU C Library is distributed in the hope that it will be useful,
>> +   but WITHOUT ANY WARRANTY; without even the implied warranty of
>> +   MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the GNU
>> +   Lesser General Public License for more details.
>> +
>> +   You should have received a copy of the GNU Lesser General Public
>> +   License along with the GNU C Library; if not, see
>> +<https://www.gnu.org/licenses/>.  */
>> +
>> +#include <sysdep.h>
>> +#define _ERRNO_H  1
>> +#include <bits/errno.h>
>> +
>> +/* Clone the calling process, but without copying the whole address space.
>> +   The calling process is suspended until the new process exits or is
>> +   replaced by a call to `execve'.  Return -1 for errors, 0 to the new process,
>> +   and the process ID of the new process to the old process.  */
>> +
>> +ENTRY (__vfork)
>> +
>> +	li.d	a0, 0x4111 /* CLONE_VM | CLONE_VFORK | SIGCHLD */
>> +	add.d   a1, zero, sp
>> +
>> +	/* Do the system call.  */
>> +	li.d	a7, __NR_clone
>> +	syscall	0
>> +
>> +	blt	a0, zero ,L (error)
>> +
>> +	ret
>> +
>> +L (error):
>> +	b	__syscall_error
>> +
>> +END (__vfork)
>> +
>> +libc_hidden_def (__vfork)
>> +
>> +weak_alias (__vfork, vfork)
>> +strong_alias (__vfork, __libc_vfork)

  reply	other threads:[~2023-09-13 11:07 UTC|newest]

Thread overview: 40+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2022-07-08  6:52 [PATCH v6 00/13] GLIBC LoongArch PATCHES caiyinyu
2022-07-08  6:52 ` [PATCH v6 01/13] LoongArch: Update NEWS and README for the LoongArch port caiyinyu
2022-07-12 11:06   ` Adhemerval Zanella Netto
2022-07-08  6:52 ` [PATCH v6 02/13] LoongArch: Add LoongArch entries to config.h.in caiyinyu
2022-07-12 18:51   ` Adhemerval Zanella Netto
2022-07-08  6:52 ` [PATCH v6 03/13] LoongArch: Add relocations and ELF flags to elf.h and scripts/glibcelf.py caiyinyu
2022-07-12 19:27   ` Adhemerval Zanella Netto
2022-07-08  6:52 ` [PATCH v6 04/13] LoongArch: ABI Implementation caiyinyu
2022-07-12 20:39   ` Adhemerval Zanella Netto
2022-07-15  1:46     ` caiyinyu
2022-07-08  6:52 ` [PATCH v6 05/13] LoongArch: Thread-Local Storage Support caiyinyu
2022-07-13 13:10   ` Adhemerval Zanella Netto
2022-07-08  6:52 ` [PATCH v6 06/13] LoongArch: Generic <math.h> and soft-fp Routines caiyinyu
2022-07-13 13:25   ` Adhemerval Zanella Netto
2022-07-08  6:52 ` [PATCH v6 07/13] LoongArch: Atomic and Locking Routines caiyinyu
2022-07-13 13:53   ` Adhemerval Zanella Netto
2022-07-15  1:46     ` caiyinyu
2022-07-08  6:52 ` [PATCH v6 08/13] LoongArch: Linux Syscall Interface caiyinyu
2022-07-13 16:19   ` Adhemerval Zanella Netto
2023-09-13 11:05     ` caiyinyu [this message]
2022-07-08  6:52 ` [PATCH v6 09/13] LoongArch: Linux ABI caiyinyu
2022-07-13 16:35   ` Adhemerval Zanella Netto
2022-07-15  1:48     ` caiyinyu
2022-07-08  6:52 ` [PATCH v6 10/13] LoongArch: Add ABI Lists caiyinyu
2022-07-13 17:12   ` Adhemerval Zanella Netto
2022-07-08  6:52 ` [PATCH v6 11/13] LoongArch: Build Infastructure caiyinyu
2022-07-13  3:07   ` Xi Ruoyao
2022-07-13  3:43     ` WANG Xuerui
2022-07-13  7:51       ` caiyinyu
2022-07-13 17:48   ` Adhemerval Zanella Netto
2022-07-15  1:49     ` caiyinyu
2022-07-08  6:52 ` [PATCH v6 12/13] LoongArch: Hard Float Support caiyinyu
2022-07-13 19:22   ` Adhemerval Zanella Netto
2022-07-14 13:00     ` caiyinyu
2022-07-08  7:13 ` [PATCH v6 00/13] GLIBC LoongArch PATCHES Xi Ruoyao
2022-07-08  7:15   ` Xi Ruoyao
2022-07-13 19:55   ` Adhemerval Zanella Netto
2022-07-14 11:33     ` Xi Ruoyao
2022-07-14 12:11       ` Adhemerval Zanella Netto
2022-07-18 13:54 ` Carlos O'Donell

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=71e13529-9bb0-f575-d026-6214534c6de1@loongson.cn \
    --to=caiyinyu@loongson.cn \
    --cc=adhemerval.zanella@linaro.org \
    --cc=joseph_myers@mentor.com \
    --cc=libc-alpha@sourceware.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).