mailing list of musl libc
 help / color / mirror / code / Atom feed
* [musl] loongarch64 merge
@ 2024-01-25 17:43 Rich Felker
  2024-01-26  1:18 ` Fangrui Song
  2024-01-29 16:47 ` Rich Felker
  0 siblings, 2 replies; 7+ messages in thread
From: Rich Felker @ 2024-01-25 17:43 UTC (permalink / raw)
  To: musl

[-- Attachment #1: Type: text/plain, Size: 1704 bytes --]

I'm going through where everything was left on this topic and
preparing a patch for merge. This message/thread is to document what
I'm actually doing vs the various submitted versions of the patch
since v5/v6 where the major review took place.


Subsequent changes I'm reverting:

- De-optimization of __get_tp. No motivation for removing the
  potentially in-place $tp was provided, and we generally use the
  arch's tp in-place unless there's a compiler bug to be worked
  around. See powerpc{,64} for an example where it's used, or1k where
  we have a probably-obsolete workaround for ancient clang being
  broken.

- unsigned -> unsigned int, etc.

- Gratuitous whitespace changes in headers that obscure the fact that
  a header is a complete duplicate that could eventually be shared
  between archs (e.g. bits/float.h, bits/posix.h) or just obscure
  what differs from other archs when running diff.


Fixes from previous review that were overlooked:

- Removing SA_RESTORER -- its presence defined as 0 produces wrong
  sigaction ABI.


Additions:

- Adding the reloc.h/configure case for single-only float.

- The new member names for mcontext_t are all in reserved namespace,
  so there's no reason to have a separate namespace-clean version of
  mcontext_t, and I'm removing the latter.

- Public member uc_flags with no __, macro for compat with any
  existing software using the __-prefixed name.


Still TODO:

I don't think I ever reviewed the apparent rewrite of sigsetjmp and
possibly some other asm that changed between v5 and v8. I'm about to
start looking at that and will follow up.


Attached are a "differences vs v8" patch and what my cumulative patch
looks like right now.

Rich

[-- Attachment #2: proposed_vs_v8.diff --]
[-- Type: text/plain, Size: 8206 bytes --]

diff --git a/arch/loongarch64/bits/alltypes.h.in b/arch/loongarch64/bits/alltypes.h.in
index 06db4096..d1807aca 100644
--- a/arch/loongarch64/bits/alltypes.h.in
+++ b/arch/loongarch64/bits/alltypes.h.in
@@ -1,15 +1,15 @@
-#define _Addr  long
+#define _Addr long
 #define _Int64 long
-#define _Reg   long
+#define _Reg long
 
-#define __BYTE_ORDER  1234
-#define __LONG_MAX    0x7fffffffffffffffL
+#define __BYTE_ORDER 1234
+#define __LONG_MAX 0x7fffffffffffffffL
 
 #ifndef __cplusplus
 TYPEDEF int wchar_t;
 #endif
 
-TYPEDEF float  float_t;
+TYPEDEF float float_t;
 TYPEDEF double double_t;
 
 TYPEDEF struct { long long __ll; long double __ld; } max_align_t;
diff --git a/arch/loongarch64/bits/fenv.h b/arch/loongarch64/bits/fenv.h
index 99e916e1..264cafb5 100644
--- a/arch/loongarch64/bits/fenv.h
+++ b/arch/loongarch64/bits/fenv.h
@@ -14,7 +14,7 @@
 typedef unsigned fexcept_t;
 
 typedef struct {
-	unsigned int __cw;
+	unsigned __cw;
 } fenv_t;
 
 #define FE_DFL_ENV ((const fenv_t *) -1)
diff --git a/arch/loongarch64/bits/float.h b/arch/loongarch64/bits/float.h
index 63e86d44..719c7908 100644
--- a/arch/loongarch64/bits/float.h
+++ b/arch/loongarch64/bits/float.h
@@ -1,16 +1,16 @@
 #define FLT_EVAL_METHOD 0
 
-#define LDBL_TRUE_MIN   6.47517511943802511092443895822764655e-4966L
-#define LDBL_MIN        3.36210314311209350626267781732175260e-4932L
-#define LDBL_MAX        1.18973149535723176508575932662800702e+4932L
-#define LDBL_EPSILON    1.92592994438723585305597794258492732e-34L
+#define LDBL_TRUE_MIN 6.47517511943802511092443895822764655e-4966L
+#define LDBL_MIN 3.36210314311209350626267781732175260e-4932L
+#define LDBL_MAX 1.18973149535723176508575932662800702e+4932L
+#define LDBL_EPSILON 1.92592994438723585305597794258492732e-34L
 
-#define LDBL_MANT_DIG   113
-#define LDBL_MIN_EXP    (-16381)
-#define LDBL_MAX_EXP    16384
+#define LDBL_MANT_DIG 113
+#define LDBL_MIN_EXP (-16381)
+#define LDBL_MAX_EXP 16384
 
-#define LDBL_DIG        33
+#define LDBL_DIG 33
 #define LDBL_MIN_10_EXP (-4931)
 #define LDBL_MAX_10_EXP 4932
 
-#define DECIMAL_DIG     36
+#define DECIMAL_DIG 36
diff --git a/arch/loongarch64/bits/posix.h b/arch/loongarch64/bits/posix.h
index c37b94c1..8068ce98 100644
--- a/arch/loongarch64/bits/posix.h
+++ b/arch/loongarch64/bits/posix.h
@@ -1,2 +1,2 @@
-#define _POSIX_V6_LP64_OFF64  1
-#define _POSIX_V7_LP64_OFF64  1
+#define _POSIX_V6_LP64_OFF64 1
+#define _POSIX_V7_LP64_OFF64 1
diff --git a/arch/loongarch64/bits/setjmp.h b/arch/loongarch64/bits/setjmp.h
index 4bfa374d..3b15e87b 100644
--- a/arch/loongarch64/bits/setjmp.h
+++ b/arch/loongarch64/bits/setjmp.h
@@ -1 +1 @@
-typedef unsigned long  __jmp_buf[23];
+typedef unsigned long __jmp_buf[23];
diff --git a/arch/loongarch64/bits/signal.h b/arch/loongarch64/bits/signal.h
index e1d256e7..8e0e605c 100644
--- a/arch/loongarch64/bits/signal.h
+++ b/arch/loongarch64/bits/signal.h
@@ -3,7 +3,7 @@
 
 #if defined(_XOPEN_SOURCE) || defined(_GNU_SOURCE) || defined(_BSD_SOURCE)
 #define MINSIGSTKSZ 4096
-#define SIGSTKSZ    16384
+#define SIGSTKSZ 16384
 #endif
 
 #if defined(_GNU_SOURCE) || defined(_BSD_SOURCE)
@@ -12,81 +12,79 @@ typedef unsigned long greg_t, gregset_t[32];
 struct sigcontext {
 	unsigned long sc_pc;
 	unsigned long sc_regs[32];
-	unsigned int sc_flags;
+	unsigned sc_flags;
 	unsigned long sc_extcontext[] __attribute__((__aligned__(16)));
 };
+#endif
 
 typedef struct {
 	unsigned long __pc;
 	unsigned long __gregs[32];
-	unsigned int __flags;
+	unsigned __flags;
 	unsigned long __extcontext[] __attribute__((__aligned__(16)));
 } mcontext_t;
-#else
-typedef struct {
-	unsigned long __space[34];
-} mcontext_t;
-#endif
 
 struct sigaltstack {
-	void   *ss_sp;
-	int    ss_flags;
+	void *ss_sp;
+	int ss_flags;
 	size_t ss_size;
 };
 
 typedef struct __ucontext
 {
-	unsigned long      __uc_flags;
-	struct __ucontext  *uc_link;
-	stack_t            uc_stack;
-	sigset_t           uc_sigmask;
-	long               __uc_pad;
-	mcontext_t         uc_mcontext;
+	unsigned long uc_flags;
+	struct __ucontext *uc_link;
+	stack_t uc_stack;
+	sigset_t uc_sigmask;
+	long __uc_pad;
+	mcontext_t uc_mcontext;
 } ucontext_t;
 
-#define SA_NOCLDSTOP     1
-#define SA_NOCLDWAIT     2
-#define SA_SIGINFO       4
-#define SA_ONSTACK       0x08000000
-#define SA_RESTART       0x10000000
-#define SA_NODEFER       0x40000000
-#define SA_RESETHAND     0x80000000
-#define SA_RESTORER      0x0
+#define __uc_flags uc_flags
+
+#define SA_NOCLDSTOP 1
+#define SA_NOCLDWAIT 2
+#define SA_SIGINFO   4
+#define SA_ONSTACK   0x08000000
+#define SA_RESTART   0x10000000
+#define SA_NODEFER   0x40000000
+#define SA_RESETHAND 0x80000000
 
 #endif
 
-#define SIGHUP           1
-#define SIGINT           2
-#define SIGQUIT          3
-#define SIGILL           4
-#define SIGTRAP          5
-#define SIGABRT          6
-#define SIGIOT           SIGABRT
-#define SIGBUS           7
-#define SIGFPE           8
-#define SIGKILL          9
-#define SIGUSR1         10
-#define SIGSEGV         11
-#define SIGUSR2         12
-#define SIGPIPE         13
-#define SIGALRM         14
-#define SIGTERM         15
-#define SIGSTKFLT       16
-#define SIGCHLD         17
-#define SIGCONT         18
-#define SIGSTOP         19
-#define SIGTSTP         20
-#define SIGTTIN         21
-#define SIGTTOU         22
-#define SIGURG          23
-#define SIGXCPU         24
-#define SIGXFSZ         25
-#define SIGVTALRM       26
-#define SIGPROF         27
-#define SIGWINCH        28
-#define SIGIO           29
-#define SIGPOLL         SIGIO
-#define SIGPWR          30
-#define SIGSYS          31
-#define SIGUNUSED       SIGSYS
-#define _NSIG           65
+#define SIGHUP     1
+#define SIGINT     2
+#define SIGQUIT    3
+#define SIGILL     4
+#define SIGTRAP    5
+#define SIGABRT    6
+#define SIGIOT     SIGABRT
+#define SIGBUS     7
+#define SIGFPE     8
+#define SIGKILL    9
+#define SIGUSR1   10
+#define SIGSEGV   11
+#define SIGUSR2   12
+#define SIGPIPE   13
+#define SIGALRM   14
+#define SIGTERM   15
+#define SIGSTKFLT 16
+#define SIGCHLD   17
+#define SIGCONT   18
+#define SIGSTOP   19
+#define SIGTSTP   20
+#define SIGTTIN   21
+#define SIGTTOU   22
+#define SIGURG    23
+#define SIGXCPU   24
+#define SIGXFSZ   25
+#define SIGVTALRM 26
+#define SIGPROF   27
+#define SIGWINCH  28
+#define SIGIO     29
+#define SIGPOLL   SIGIO
+#define SIGPWR    30
+#define SIGSYS    31
+#define SIGUNUSED SIGSYS
+
+#define _NSIG 65
diff --git a/arch/loongarch64/bits/stdint.h b/arch/loongarch64/bits/stdint.h
index 60c12499..1bb147f2 100644
--- a/arch/loongarch64/bits/stdint.h
+++ b/arch/loongarch64/bits/stdint.h
@@ -1,5 +1,5 @@
-typedef int32_t  int_fast16_t;
-typedef int32_t  int_fast32_t;
+typedef int32_t int_fast16_t;
+typedef int32_t int_fast32_t;
 typedef uint32_t uint_fast16_t;
 typedef uint32_t uint_fast32_t;
 
diff --git a/arch/loongarch64/pthread_arch.h b/arch/loongarch64/pthread_arch.h
index 28fbfcd1..365f6ca8 100644
--- a/arch/loongarch64/pthread_arch.h
+++ b/arch/loongarch64/pthread_arch.h
@@ -1,7 +1,7 @@
 static inline uintptr_t __get_tp()
 {
-	uintptr_t tp;
-	__asm__ __volatile__("move %0, $tp" : "=r"(tp));
+	register uintptr_t tp __asm__("tp");
+	__asm__ ("" : "=r" (tp) );
 	return tp;
 }
 
diff --git a/arch/loongarch64/reloc.h b/arch/loongarch64/reloc.h
index a4482b48..61eaca9e 100644
--- a/arch/loongarch64/reloc.h
+++ b/arch/loongarch64/reloc.h
@@ -1,10 +1,12 @@
 #ifdef __loongarch_soft_float
 #define FP_SUFFIX "-sf"
+#elif defined __loongarch_single_float
+#define FP_SUFFIX "-sp"
 #else
 #define FP_SUFFIX ""
 #endif
 
-#define LDSO_ARCH "loongarch64"  FP_SUFFIX
+#define LDSO_ARCH "loongarch64" FP_SUFFIX
 
 #define TPOFF_K         0
 
diff --git a/configure b/configure
index 55d179f1..940b3eaf 100755
--- a/configure
+++ b/configure
@@ -674,7 +674,8 @@ fi
 
 if test "$ARCH" = "loongarch64" ; then
 trycppif __loongarch_soft_float "$t" && SUBARCH=${SUBARCH}-sf
-printf "checking whether compiler support FCSRs... "
+trycppif __loongarch_single_float "$t" && SUBARCH=${SUBARCH}-sp
+printf "checking whether assembler support FCSRs... "
 echo "__asm__(\"movfcsr2gr \$t0,\$fcsr0\");" > "$tmpc"
 if $CC -c -o /dev/null "$tmpc" >/dev/null 2>&1 ; then
 printf "yes\n"

[-- Attachment #3: proposed_cumulative.diff --]
[-- Type: text/plain, Size: 41273 bytes --]

diff --git a/arch/loongarch64/atomic_arch.h b/arch/loongarch64/atomic_arch.h
new file mode 100644
index 00000000..2225d027
--- /dev/null
+++ b/arch/loongarch64/atomic_arch.h
@@ -0,0 +1,53 @@
+#define a_ll a_ll
+static inline int a_ll(volatile int *p)
+{
+	int v;
+	__asm__ __volatile__ (
+		"ll.w %0, %1"
+		: "=r"(v)
+		: "ZC"(*p));
+	return v;
+}
+
+#define a_sc a_sc
+static inline int a_sc(volatile int *p, int v)
+{
+	int r;
+	__asm__ __volatile__ (
+		"sc.w %0, %1"
+		: "=r"(r), "=ZC"(*p)
+		: "0"(v) : "memory");
+	return r;
+}
+
+#define a_ll_p a_ll_p
+static inline void *a_ll_p(volatile void *p)
+{
+	void *v;
+	__asm__ __volatile__ (
+		"ll.d %0, %1"
+		: "=r"(v)
+		: "ZC"(*(void *volatile *)p));
+	return v;
+}
+
+#define a_sc_p a_sc_p
+static inline int a_sc_p(volatile void *p, void *v)
+{
+	long r;
+	__asm__ __volatile__ (
+		"sc.d %0, %1"
+		: "=r"(r), "=ZC"(*(void *volatile *)p)
+		: "0"(v)
+		: "memory");
+	return r;
+}
+
+#define a_barrier a_barrier
+static inline void a_barrier()
+{
+	__asm__ __volatile__ ("dbar 0" : : : "memory");
+}
+
+#define a_pre_llsc  a_barrier
+#define a_post_llsc a_barrier
diff --git a/arch/loongarch64/bits/alltypes.h.in b/arch/loongarch64/bits/alltypes.h.in
new file mode 100644
index 00000000..06db4096
--- /dev/null
+++ b/arch/loongarch64/bits/alltypes.h.in
@@ -0,0 +1,18 @@
+#define _Addr  long
+#define _Int64 long
+#define _Reg   long
+
+#define __BYTE_ORDER  1234
+#define __LONG_MAX    0x7fffffffffffffffL
+
+#ifndef __cplusplus
+TYPEDEF int wchar_t;
+#endif
+
+TYPEDEF float  float_t;
+TYPEDEF double double_t;
+
+TYPEDEF struct { long long __ll; long double __ld; } max_align_t;
+
+TYPEDEF unsigned nlink_t;
+TYPEDEF int blksize_t;
diff --git a/arch/loongarch64/bits/fenv.h b/arch/loongarch64/bits/fenv.h
new file mode 100644
index 00000000..99e916e1
--- /dev/null
+++ b/arch/loongarch64/bits/fenv.h
@@ -0,0 +1,20 @@
+#define FE_INEXACT    0x010000
+#define FE_UNDERFLOW  0x020000
+#define FE_OVERFLOW   0x040000
+#define FE_DIVBYZERO  0x080000
+#define FE_INVALID    0x100000
+
+#define FE_ALL_EXCEPT 0x1F0000
+
+#define FE_TONEAREST  0x000
+#define FE_TOWARDZERO 0x100
+#define FE_UPWARD     0x200
+#define FE_DOWNWARD   0x300
+
+typedef unsigned fexcept_t;
+
+typedef struct {
+	unsigned int __cw;
+} fenv_t;
+
+#define FE_DFL_ENV ((const fenv_t *) -1)
diff --git a/arch/loongarch64/bits/float.h b/arch/loongarch64/bits/float.h
new file mode 100644
index 00000000..63e86d44
--- /dev/null
+++ b/arch/loongarch64/bits/float.h
@@ -0,0 +1,16 @@
+#define FLT_EVAL_METHOD 0
+
+#define LDBL_TRUE_MIN   6.47517511943802511092443895822764655e-4966L
+#define LDBL_MIN        3.36210314311209350626267781732175260e-4932L
+#define LDBL_MAX        1.18973149535723176508575932662800702e+4932L
+#define LDBL_EPSILON    1.92592994438723585305597794258492732e-34L
+
+#define LDBL_MANT_DIG   113
+#define LDBL_MIN_EXP    (-16381)
+#define LDBL_MAX_EXP    16384
+
+#define LDBL_DIG        33
+#define LDBL_MIN_10_EXP (-4931)
+#define LDBL_MAX_10_EXP 4932
+
+#define DECIMAL_DIG     36
diff --git a/arch/loongarch64/bits/posix.h b/arch/loongarch64/bits/posix.h
new file mode 100644
index 00000000..c37b94c1
--- /dev/null
+++ b/arch/loongarch64/bits/posix.h
@@ -0,0 +1,2 @@
+#define _POSIX_V6_LP64_OFF64  1
+#define _POSIX_V7_LP64_OFF64  1
diff --git a/arch/loongarch64/bits/ptrace.h b/arch/loongarch64/bits/ptrace.h
new file mode 100644
index 00000000..dce2fa51
--- /dev/null
+++ b/arch/loongarch64/bits/ptrace.h
@@ -0,0 +1,4 @@
+#define PTRACE_GET_THREAD_AREA  25
+#define PTRACE_SET_THREAD_AREA  26
+#define PTRACE_GET_WATCH_REGS   0xd0
+#define PTRACE_SET_WATCH_REGS   0xd1
diff --git a/arch/loongarch64/bits/reg.h b/arch/loongarch64/bits/reg.h
new file mode 100644
index 00000000..2633f39d
--- /dev/null
+++ b/arch/loongarch64/bits/reg.h
@@ -0,0 +1,2 @@
+#undef __WORDSIZE
+#define __WORDSIZE 64
diff --git a/arch/loongarch64/bits/setjmp.h b/arch/loongarch64/bits/setjmp.h
new file mode 100644
index 00000000..4bfa374d
--- /dev/null
+++ b/arch/loongarch64/bits/setjmp.h
@@ -0,0 +1 @@
+typedef unsigned long  __jmp_buf[23];
diff --git a/arch/loongarch64/bits/signal.h b/arch/loongarch64/bits/signal.h
new file mode 100644
index 00000000..e1d256e7
--- /dev/null
+++ b/arch/loongarch64/bits/signal.h
@@ -0,0 +1,92 @@
+#if defined(_POSIX_SOURCE) || defined(_POSIX_C_SOURCE) \
+ || defined(_XOPEN_SOURCE) || defined(_GNU_SOURCE) || defined(_BSD_SOURCE)
+
+#if defined(_XOPEN_SOURCE) || defined(_GNU_SOURCE) || defined(_BSD_SOURCE)
+#define MINSIGSTKSZ 4096
+#define SIGSTKSZ    16384
+#endif
+
+#if defined(_GNU_SOURCE) || defined(_BSD_SOURCE)
+typedef unsigned long greg_t, gregset_t[32];
+
+struct sigcontext {
+	unsigned long sc_pc;
+	unsigned long sc_regs[32];
+	unsigned int sc_flags;
+	unsigned long sc_extcontext[] __attribute__((__aligned__(16)));
+};
+
+typedef struct {
+	unsigned long __pc;
+	unsigned long __gregs[32];
+	unsigned int __flags;
+	unsigned long __extcontext[] __attribute__((__aligned__(16)));
+} mcontext_t;
+#else
+typedef struct {
+	unsigned long __space[34];
+} mcontext_t;
+#endif
+
+struct sigaltstack {
+	void   *ss_sp;
+	int    ss_flags;
+	size_t ss_size;
+};
+
+typedef struct __ucontext
+{
+	unsigned long      __uc_flags;
+	struct __ucontext  *uc_link;
+	stack_t            uc_stack;
+	sigset_t           uc_sigmask;
+	long               __uc_pad;
+	mcontext_t         uc_mcontext;
+} ucontext_t;
+
+#define SA_NOCLDSTOP     1
+#define SA_NOCLDWAIT     2
+#define SA_SIGINFO       4
+#define SA_ONSTACK       0x08000000
+#define SA_RESTART       0x10000000
+#define SA_NODEFER       0x40000000
+#define SA_RESETHAND     0x80000000
+#define SA_RESTORER      0x0
+
+#endif
+
+#define SIGHUP           1
+#define SIGINT           2
+#define SIGQUIT          3
+#define SIGILL           4
+#define SIGTRAP          5
+#define SIGABRT          6
+#define SIGIOT           SIGABRT
+#define SIGBUS           7
+#define SIGFPE           8
+#define SIGKILL          9
+#define SIGUSR1         10
+#define SIGSEGV         11
+#define SIGUSR2         12
+#define SIGPIPE         13
+#define SIGALRM         14
+#define SIGTERM         15
+#define SIGSTKFLT       16
+#define SIGCHLD         17
+#define SIGCONT         18
+#define SIGSTOP         19
+#define SIGTSTP         20
+#define SIGTTIN         21
+#define SIGTTOU         22
+#define SIGURG          23
+#define SIGXCPU         24
+#define SIGXFSZ         25
+#define SIGVTALRM       26
+#define SIGPROF         27
+#define SIGWINCH        28
+#define SIGIO           29
+#define SIGPOLL         SIGIO
+#define SIGPWR          30
+#define SIGSYS          31
+#define SIGUNUSED       SIGSYS
+#define _NSIG           65
diff --git a/arch/loongarch64/bits/stat.h b/arch/loongarch64/bits/stat.h
new file mode 100644
index 00000000..b7f4221b
--- /dev/null
+++ b/arch/loongarch64/bits/stat.h
@@ -0,0 +1,18 @@
+struct stat {
+	dev_t st_dev;
+	ino_t st_ino;
+	mode_t st_mode;
+	nlink_t st_nlink;
+	uid_t st_uid;
+	gid_t st_gid;
+	dev_t st_rdev;
+	unsigned long __pad;
+	off_t st_size;
+	blksize_t st_blksize;
+	int __pad2;
+	blkcnt_t st_blocks;
+	struct timespec st_atim;
+	struct timespec st_mtim;
+	struct timespec st_ctim;
+	unsigned __unused[2];
+};
diff --git a/arch/loongarch64/bits/stdint.h b/arch/loongarch64/bits/stdint.h
new file mode 100644
index 00000000..60c12499
--- /dev/null
+++ b/arch/loongarch64/bits/stdint.h
@@ -0,0 +1,20 @@
+typedef int32_t  int_fast16_t;
+typedef int32_t  int_fast32_t;
+typedef uint32_t uint_fast16_t;
+typedef uint32_t uint_fast32_t;
+
+#define INT_FAST16_MIN  INT32_MIN
+#define INT_FAST32_MIN  INT32_MIN
+
+#define INT_FAST16_MAX  INT32_MAX
+#define INT_FAST32_MAX  INT32_MAX
+
+#define UINT_FAST16_MAX UINT32_MAX
+#define UINT_FAST32_MAX UINT32_MAX
+
+#define INTPTR_MIN      INT64_MIN
+#define INTPTR_MAX      INT64_MAX
+#define UINTPTR_MAX     UINT64_MAX
+#define PTRDIFF_MIN     INT64_MIN
+#define PTRDIFF_MAX     INT64_MAX
+#define SIZE_MAX        UINT64_MAX
diff --git a/arch/loongarch64/bits/syscall.h.in b/arch/loongarch64/bits/syscall.h.in
new file mode 100644
index 00000000..0980e533
--- /dev/null
+++ b/arch/loongarch64/bits/syscall.h.in
@@ -0,0 +1,303 @@
+#define __NR_io_setup                   0
+#define __NR_io_destroy                 1
+#define __NR_io_submit                  2
+#define __NR_io_cancel                  3
+#define __NR_io_getevents               4
+#define __NR_setxattr                   5
+#define __NR_lsetxattr                  6
+#define __NR_fsetxattr                  7
+#define __NR_getxattr                   8
+#define __NR_lgetxattr                  9
+#define __NR_fgetxattr                  10
+#define __NR_listxattr                  11
+#define __NR_llistxattr                 12
+#define __NR_flistxattr                 13
+#define __NR_removexattr                14
+#define __NR_lremovexattr               15
+#define __NR_fremovexattr               16
+#define __NR_getcwd                     17
+#define __NR_lookup_dcookie             18
+#define __NR_eventfd2                   19
+#define __NR_epoll_create1              20
+#define __NR_epoll_ctl                  21
+#define __NR_epoll_pwait                22
+#define __NR_dup                        23
+#define __NR_dup3                       24
+#define __NR3264_fcntl                  25
+#define __NR_inotify_init1              26
+#define __NR_inotify_add_watch          27
+#define __NR_inotify_rm_watch           28
+#define __NR_ioctl                      29
+#define __NR_ioprio_set                 30
+#define __NR_ioprio_get                 31
+#define __NR_flock                      32
+#define __NR_mknodat                    33
+#define __NR_mkdirat                    34
+#define __NR_unlinkat                   35
+#define __NR_symlinkat                  36
+#define __NR_linkat                     37
+#define __NR_umount2                    39
+#define __NR_mount                      40
+#define __NR_pivot_root                 41
+#define __NR_nfsservctl                 42
+#define __NR3264_statfs                 43
+#define __NR3264_fstatfs                44
+#define __NR3264_truncate               45
+#define __NR3264_ftruncate              46
+#define __NR_fallocate                  47
+#define __NR_faccessat                  48
+#define __NR_chdir                      49
+#define __NR_fchdir                     50
+#define __NR_chroot                     51
+#define __NR_fchmod                     52
+#define __NR_fchmodat                   53
+#define __NR_fchownat                   54
+#define __NR_fchown                     55
+#define __NR_openat                     56
+#define __NR_close                      57
+#define __NR_vhangup                    58
+#define __NR_pipe2                      59
+#define __NR_quotactl                   60
+#define __NR_getdents64                 61
+#define __NR3264_lseek                  62
+#define __NR_read                       63
+#define __NR_write                      64
+#define __NR_readv                      65
+#define __NR_writev                     66
+#define __NR_pread64                    67
+#define __NR_pwrite64                   68
+#define __NR_preadv                     69
+#define __NR_pwritev                    70
+#define __NR3264_sendfile               71
+#define __NR_pselect6                   72
+#define __NR_ppoll                      73
+#define __NR_signalfd4                  74
+#define __NR_vmsplice                   75
+#define __NR_splice                     76
+#define __NR_tee                        77
+#define __NR_readlinkat                 78
+#define __NR_sync                       81
+#define __NR_fsync                      82
+#define __NR_fdatasync                  83
+#define __NR_sync_file_range            84
+#define __NR_timerfd_create             85
+#define __NR_timerfd_settime            86
+#define __NR_timerfd_gettime            87
+#define __NR_utimensat                  88
+#define __NR_acct                       89
+#define __NR_capget                     90
+#define __NR_capset                     91
+#define __NR_personality                92
+#define __NR_exit                       93
+#define __NR_exit_group                 94
+#define __NR_waitid                     95
+#define __NR_set_tid_address            96
+#define __NR_unshare                    97
+#define __NR_futex                      98
+#define __NR_set_robust_list            99
+#define __NR_get_robust_list            100
+#define __NR_nanosleep                  101
+#define __NR_getitimer                  102
+#define __NR_setitimer                  103
+#define __NR_kexec_load                 104
+#define __NR_init_module                105
+#define __NR_delete_module              106
+#define __NR_timer_create               107
+#define __NR_timer_gettime              108
+#define __NR_timer_getoverrun           109
+#define __NR_timer_settime              110
+#define __NR_timer_delete               111
+#define __NR_clock_settime              112
+#define __NR_clock_gettime              113
+#define __NR_clock_getres               114
+#define __NR_clock_nanosleep            115
+#define __NR_syslog                     116
+#define __NR_ptrace                     117
+#define __NR_sched_setparam             118
+#define __NR_sched_setscheduler         119
+#define __NR_sched_getscheduler         120
+#define __NR_sched_getparam             121
+#define __NR_sched_setaffinity          122
+#define __NR_sched_getaffinity          123
+#define __NR_sched_yield                124
+#define __NR_sched_get_priority_max     125
+#define __NR_sched_get_priority_min     126
+#define __NR_sched_rr_get_interval      127
+#define __NR_restart_syscall            128
+#define __NR_kill                       129
+#define __NR_tkill                      130
+#define __NR_tgkill                     131
+#define __NR_sigaltstack                132
+#define __NR_rt_sigsuspend              133
+#define __NR_rt_sigaction               134
+#define __NR_rt_sigprocmask             135
+#define __NR_rt_sigpending              136
+#define __NR_rt_sigtimedwait            137
+#define __NR_rt_sigqueueinfo            138
+#define __NR_rt_sigreturn               139
+#define __NR_setpriority                140
+#define __NR_getpriority                141
+#define __NR_reboot                     142
+#define __NR_setregid                   143
+#define __NR_setgid                     144
+#define __NR_setreuid                   145
+#define __NR_setuid                     146
+#define __NR_setresuid                  147
+#define __NR_getresuid                  148
+#define __NR_setresgid                  149
+#define __NR_getresgid                  150
+#define __NR_setfsuid                   151
+#define __NR_setfsgid                   152
+#define __NR_times                      153
+#define __NR_setpgid                    154
+#define __NR_getpgid                    155
+#define __NR_getsid                     156
+#define __NR_setsid                     157
+#define __NR_getgroups                  158
+#define __NR_setgroups                  159
+#define __NR_uname                      160
+#define __NR_sethostname                161
+#define __NR_setdomainname              162
+#define __NR_getrlimit                  163
+#define __NR_setrlimit                  164
+#define __NR_getrusage                  165
+#define __NR_umask                      166
+#define __NR_prctl                      167
+#define __NR_getcpu                     168
+#define __NR_gettimeofday               169
+#define __NR_settimeofday               170
+#define __NR_adjtimex                   171
+#define __NR_getpid                     172
+#define __NR_getppid                    173
+#define __NR_getuid                     174
+#define __NR_geteuid                    175
+#define __NR_getgid                     176
+#define __NR_getegid                    177
+#define __NR_gettid                     178
+#define __NR_sysinfo                    179
+#define __NR_mq_open                    180
+#define __NR_mq_unlink                  181
+#define __NR_mq_timedsend               182
+#define __NR_mq_timedreceive            183
+#define __NR_mq_notify                  184
+#define __NR_mq_getsetattr              185
+#define __NR_msgget                     186
+#define __NR_msgctl                     187
+#define __NR_msgrcv                     188
+#define __NR_msgsnd                     189
+#define __NR_semget                     190
+#define __NR_semctl                     191
+#define __NR_semtimedop                 192
+#define __NR_semop                      193
+#define __NR_shmget                     194
+#define __NR_shmctl                     195
+#define __NR_shmat                      196
+#define __NR_shmdt                      197
+#define __NR_socket                     198
+#define __NR_socketpair                 199
+#define __NR_bind                       200
+#define __NR_listen                     201
+#define __NR_accept                     202
+#define __NR_connect                    203
+#define __NR_getsockname                204
+#define __NR_getpeername                205
+#define __NR_sendto                     206
+#define __NR_recvfrom                   207
+#define __NR_setsockopt                 208
+#define __NR_getsockopt                 209
+#define __NR_shutdown                   210
+#define __NR_sendmsg                    211
+#define __NR_recvmsg                    212
+#define __NR_readahead                  213
+#define __NR_brk                        214
+#define __NR_munmap                     215
+#define __NR_mremap                     216
+#define __NR_add_key                    217
+#define __NR_request_key                218
+#define __NR_keyctl                     219
+#define __NR_clone                      220
+#define __NR_execve                     221
+#define __NR3264_mmap                   222
+#define __NR3264_fadvise64              223
+#define __NR_swapon                     224
+#define __NR_swapoff                    225
+#define __NR_mprotect                   226
+#define __NR_msync                      227
+#define __NR_mlock                      228
+#define __NR_munlock                    229
+#define __NR_mlockall                   230
+#define __NR_munlockall                 231
+#define __NR_mincore                    232
+#define __NR_madvise                    233
+#define __NR_remap_file_pages           234
+#define __NR_mbind                      235
+#define __NR_get_mempolicy              236
+#define __NR_set_mempolicy              237
+#define __NR_migrate_pages              238
+#define __NR_move_pages                 239
+#define __NR_rt_tgsigqueueinfo          240
+#define __NR_perf_event_open            241
+#define __NR_accept4                    242
+#define __NR_recvmmsg                   243
+#define __NR_arch_specific_syscall      244
+#define __NR_wait4                      260
+#define __NR_prlimit64                  261
+#define __NR_fanotify_init              262
+#define __NR_fanotify_mark              263
+#define __NR_name_to_handle_at          264
+#define __NR_open_by_handle_at          265
+#define __NR_clock_adjtime              266
+#define __NR_syncfs                     267
+#define __NR_setns                      268
+#define __NR_sendmmsg                   269
+#define __NR_process_vm_readv           270
+#define __NR_process_vm_writev          271
+#define __NR_kcmp                       272
+#define __NR_finit_module               273
+#define __NR_sched_setattr              274
+#define __NR_sched_getattr              275
+#define __NR_renameat2                  276
+#define __NR_seccomp                    277
+#define __NR_getrandom                  278
+#define __NR_memfd_create               279
+#define __NR_bpf                        280
+#define __NR_execveat                   281
+#define __NR_userfaultfd                282
+#define __NR_membarrier                 283
+#define __NR_mlock2                     284
+#define __NR_copy_file_range            285
+#define __NR_preadv2                    286
+#define __NR_pwritev2                   287
+#define __NR_pkey_mprotect              288
+#define __NR_pkey_alloc                 289
+#define __NR_pkey_free                  290
+#define __NR_statx                      291
+#define __NR_io_pgetevents              292
+#define __NR_rseq                       293
+#define __NR_kexec_file_load            294
+#define __NR_pidfd_send_signal          424
+#define __NR_io_uring_setup             425
+#define __NR_io_uring_enter             426
+#define __NR_io_uring_register          427
+#define __NR_open_tree                  428
+#define __NR_move_mount                 429
+#define __NR_fsopen                     430
+#define __NR_fsconfig                   431
+#define __NR_fsmount                    432
+#define __NR_fspick                     433
+#define __NR_pidfd_open                 434
+#define __NR_clone3                     435
+#define __NR_close_range                436
+#define __NR_openat2                    437
+#define __NR_pidfd_getfd                438
+#define __NR_faccessat2                 439
+#define __NR_process_madvise            440
+#define __NR_fcntl                      __NR3264_fcntl
+#define __NR_statfs                     __NR3264_statfs
+#define __NR_fstatfs                    __NR3264_fstatfs
+#define __NR_truncate                   __NR3264_truncate
+#define __NR_ftruncate                  __NR3264_ftruncate
+#define __NR_lseek                      __NR3264_lseek
+#define __NR_sendfile                   __NR3264_sendfile
+#define __NR_mmap                       __NR3264_mmap
+#define __NR_fadvise64                  __NR3264_fadvise64
diff --git a/arch/loongarch64/bits/user.h b/arch/loongarch64/bits/user.h
new file mode 100644
index 00000000..5a71d132
--- /dev/null
+++ b/arch/loongarch64/bits/user.h
@@ -0,0 +1,5 @@
+#define ELF_NGREG    45
+#define ELF_NFPREG   33
+
+typedef unsigned long elf_greg_t, elf_gregset_t[ELF_NGREG];
+typedef double elf_fpreg_t, elf_fpregset_t[ELF_NFPREG];
diff --git a/arch/loongarch64/crt_arch.h b/arch/loongarch64/crt_arch.h
new file mode 100644
index 00000000..e0760d9e
--- /dev/null
+++ b/arch/loongarch64/crt_arch.h
@@ -0,0 +1,13 @@
+__asm__(
+".text \n"
+".global " START "\n"
+".type   " START ", @function\n"
+START ":\n"
+"	move $fp, $zero\n"
+"	move $a0, $sp\n"
+".weak _DYNAMIC\n"
+".hidden _DYNAMIC\n"
+"	la.local $a1, _DYNAMIC\n"
+"	bstrins.d $sp, $zero, 3, 0\n"
+"	b " START "_c\n"
+);
diff --git a/arch/loongarch64/pthread_arch.h b/arch/loongarch64/pthread_arch.h
new file mode 100644
index 00000000..28fbfcd1
--- /dev/null
+++ b/arch/loongarch64/pthread_arch.h
@@ -0,0 +1,11 @@
+static inline uintptr_t __get_tp()
+{
+	uintptr_t tp;
+	__asm__ __volatile__("move %0, $tp" : "=r"(tp));
+	return tp;
+}
+
+#define TLS_ABOVE_TP
+#define GAP_ABOVE_TP   0
+#define DTP_OFFSET     0
+#define MC_PC          __pc
diff --git a/arch/loongarch64/reloc.h b/arch/loongarch64/reloc.h
new file mode 100644
index 00000000..a4482b48
--- /dev/null
+++ b/arch/loongarch64/reloc.h
@@ -0,0 +1,27 @@
+#ifdef __loongarch_soft_float
+#define FP_SUFFIX "-sf"
+#else
+#define FP_SUFFIX ""
+#endif
+
+#define LDSO_ARCH "loongarch64"  FP_SUFFIX
+
+#define TPOFF_K         0
+
+#define REL_PLT         R_LARCH_JUMP_SLOT
+#define REL_COPY        R_LARCH_COPY
+#define REL_DTPMOD      R_LARCH_TLS_DTPMOD64
+#define REL_DTPOFF      R_LARCH_TLS_DTPREL64
+#define REL_TPOFF       R_LARCH_TLS_TPREL64
+#define REL_RELATIVE    R_LARCH_RELATIVE
+#define REL_SYMBOLIC    R_LARCH_64
+
+#define CRTJMP(pc,sp) __asm__ __volatile__( \
+	"move $sp, %1 ; jr %0" : : "r"(pc), "r"(sp) : "memory" )
+
+#define GETFUNCSYM(fp, sym, got) __asm__ ( \
+	".hidden " #sym "\n" \
+	".align 8 \n" \
+	"	la.local $t1, "#sym" \n" \
+	"	move %0, $t1 \n" \
+	: "=r"(*(fp)) : : "memory" )
diff --git a/arch/loongarch64/syscall_arch.h b/arch/loongarch64/syscall_arch.h
new file mode 100644
index 00000000..4d5e1885
--- /dev/null
+++ b/arch/loongarch64/syscall_arch.h
@@ -0,0 +1,137 @@
+#define __SYSCALL_LL_E(x) (x)
+#define __SYSCALL_LL_O(x) (x)
+
+#define SYSCALL_CLOBBERLIST \
+	"$t0", "$t1", "$t2", "$t3", \
+	"$t4", "$t5", "$t6", "$t7", "$t8", "memory"
+
+static inline long __syscall0(long n)
+{
+	register long a7 __asm__("$a7") = n;
+	register long a0 __asm__("$a0");
+
+	__asm__ __volatile__ (
+		"syscall 0"
+		: "=r"(a0)
+		: "r"(a7)
+		: SYSCALL_CLOBBERLIST);
+	return a0;
+}
+
+static inline long __syscall1(long n, long a)
+{
+	register long a7 __asm__("$a7") = n;
+	register long a0 __asm__("$a0") = a;
+
+	__asm__ __volatile__ (
+		"syscall 0"
+		: "+r"(a0)
+		: "r"(a7)
+		: SYSCALL_CLOBBERLIST);
+	return a0;
+}
+
+static inline long __syscall2(long n, long a, long b)
+{
+	register long a7 __asm__("$a7") = n;
+	register long a0 __asm__("$a0") = a;
+	register long a1 __asm__("$a1") = b;
+
+	__asm__ __volatile__ (
+		"syscall 0"
+		: "+r"(a0)
+	        : "r"(a7), "r"(a1)
+		: SYSCALL_CLOBBERLIST);
+	return a0;
+}
+
+static inline long __syscall3(long n, long a, long b, long c)
+{
+	register long a7 __asm__("$a7") = n;
+	register long a0 __asm__("$a0") = a;
+	register long a1 __asm__("$a1") = b;
+	register long a2 __asm__("$a2") = c;
+
+	__asm__ __volatile__ (
+		"syscall 0"
+		: "+r"(a0)
+	        : "r"(a7), "r"(a1), "r"(a2)
+		: SYSCALL_CLOBBERLIST);
+	return a0;
+}
+
+static inline long __syscall4(long n, long a, long b, long c, long d)
+{
+	register long a7 __asm__("$a7") = n;
+	register long a0 __asm__("$a0") = a;
+	register long a1 __asm__("$a1") = b;
+	register long a2 __asm__("$a2") = c;
+	register long a3 __asm__("$a3") = d;
+
+	__asm__ __volatile__ (
+		"syscall 0"
+		: "+r"(a0)
+	        : "r"(a7), "r"(a1), "r"(a2), "r"(a3)
+		: SYSCALL_CLOBBERLIST);
+	return a0;
+}
+
+static inline long __syscall5(long n, long a, long b, long c, long d, long e)
+{
+	register long a7 __asm__("$a7") = n;
+	register long a0 __asm__("$a0") = a;
+	register long a1 __asm__("$a1") = b;
+	register long a2 __asm__("$a2") = c;
+	register long a3 __asm__("$a3") = d;
+	register long a4 __asm__("$a4") = e;
+
+	__asm__ __volatile__ (
+		"syscall 0"
+		: "+r"(a0)
+	        : "r"(a7), "r"(a1), "r"(a2), "r"(a3), "r"(a4)
+		: SYSCALL_CLOBBERLIST);
+	return a0;
+}
+
+static inline long __syscall6(long n, long a, long b, long c, long d, long e, long f)
+{
+	register long a7 __asm__("$a7") = n;
+	register long a0 __asm__("$a0") = a;
+	register long a1 __asm__("$a1") = b;
+	register long a2 __asm__("$a2") = c;
+	register long a3 __asm__("$a3") = d;
+	register long a4 __asm__("$a4") = e;
+	register long a5 __asm__("$a5") = f;
+
+	__asm__ __volatile__ (
+		"syscall 0"
+		: "+r"(a0)
+	        : "r"(a7), "r"(a1), "r"(a2), "r"(a3), "r"(a4), "r"(a5)
+		: SYSCALL_CLOBBERLIST);
+	return a0;
+}
+
+static inline long __syscall7(long n, long a, long b, long c, long d, long e, long f, long g)
+{
+	register long a7 __asm__("$a7") = n;
+	register long a0 __asm__("$a0") = a;
+	register long a1 __asm__("$a1") = b;
+	register long a2 __asm__("$a2") = c;
+	register long a3 __asm__("$a3") = d;
+	register long a4 __asm__("$a4") = e;
+	register long a5 __asm__("$a5") = f;
+	register long a6 __asm__("$a6") = g;
+
+	__asm__ __volatile__ (
+		"syscall 0"
+		: "+r"(a0)
+	        : "r"(a7), "r"(a1), "r"(a2), "r"(a3), "r"(a4), "r"(a5), "r"(a6)
+		: SYSCALL_CLOBBERLIST);
+	return a0;
+}
+
+#define VDSO_USEFUL
+#define VDSO_CGT_SYM "__vdso_clock_gettime"
+#define VDSO_CGT_VER "LINUX_5.10"
+
+#define IPC_64  0
diff --git a/configure b/configure
index 0b966ede..55d179f1 100755
--- a/configure
+++ b/configure
@@ -328,6 +328,7 @@ i?86*) ARCH=i386 ;;
 x86_64-x32*|x32*|x86_64*x32) ARCH=x32 ;;
 x86_64-nt64*) ARCH=nt64 ;;
 x86_64*) ARCH=x86_64 ;;
+loongarch64*) ARCH=loongarch64 ;;
 m68k*) ARCH=m68k ;;
 mips64*|mipsisa64*) ARCH=mips64 ;;
 mips*) ARCH=mips ;;
@@ -671,6 +672,18 @@ if test "$ARCH" = "aarch64" ; then
 trycppif __AARCH64EB__ "$t" && SUBARCH=${SUBARCH}_be
 fi
 
+if test "$ARCH" = "loongarch64" ; then
+trycppif __loongarch_soft_float "$t" && SUBARCH=${SUBARCH}-sf
+printf "checking whether compiler support FCSRs... "
+echo "__asm__(\"movfcsr2gr \$t0,\$fcsr0\");" > "$tmpc"
+if $CC -c -o /dev/null "$tmpc" >/dev/null 2>&1 ; then
+printf "yes\n"
+else
+printf "no\n"
+CFLAGS_AUTO="$CFLAGS_AUTO -DBROKEN_LOONGARCH_FCSR_ASM"
+fi
+fi
+
 if test "$ARCH" = "m68k" ; then
 if trycppif "__HAVE_68881__" ; then : ;
 elif trycppif "__mcffpu__" ; then SUBARCH="-fp64"
diff --git a/include/elf.h b/include/elf.h
index 72d17c3a..1ee72aab 100644
--- a/include/elf.h
+++ b/include/elf.h
@@ -315,7 +315,8 @@ typedef struct {
 #define EM_RISCV	243
 #define EM_BPF		247
 #define EM_CSKY		252
-#define EM_NUM		253
+#define EM_LOONGARCH	258
+#define EM_NUM		259
 
 #define EM_ALPHA	0x9026
 
@@ -704,6 +705,11 @@ typedef struct {
 #define NT_MIPS_FP_MODE	0x801
 #define NT_MIPS_MSA	0x802
 #define NT_VERSION	1
+#define NT_LOONGARCH_CPUCFG	0xa00
+#define NT_LOONGARCH_CSR	0xa01
+#define NT_LOONGARCH_LSX	0xa02
+#define NT_LOONGARCH_LASX	0xa03
+#define NT_LOONGARCH_LBT	0xa04
 
 
 
@@ -3298,6 +3304,102 @@ enum
 #define R_RISCV_SET32           56
 #define R_RISCV_32_PCREL        57
 
+#define EF_LARCH_ABI_MODIFIER_MASK    0x07
+#define EF_LARCH_ABI_SOFT_FLOAT       0x01
+#define EF_LARCH_ABI_SINGLE_FLOAT     0x02
+#define EF_LARCH_ABI_DOUBLE_FLOAT     0x03
+#define EF_LARCH_OBJABI_V1            0x40
+
+#define R_LARCH_NONE                        0
+#define R_LARCH_32                          1
+#define R_LARCH_64                          2
+#define R_LARCH_RELATIVE                    3
+#define R_LARCH_COPY                        4
+#define R_LARCH_JUMP_SLOT                   5
+#define R_LARCH_TLS_DTPMOD32                6
+#define R_LARCH_TLS_DTPMOD64                7
+#define R_LARCH_TLS_DTPREL32                8
+#define R_LARCH_TLS_DTPREL64                9
+#define R_LARCH_TLS_TPREL32                 10
+#define R_LARCH_TLS_TPREL64                 11
+#define R_LARCH_IRELATIVE                   12
+#define R_LARCH_MARK_LA                     20
+#define R_LARCH_MARK_PCREL                  21
+#define R_LARCH_SOP_PUSH_PCREL              22
+#define R_LARCH_SOP_PUSH_ABSOLUTE           23
+#define R_LARCH_SOP_PUSH_DUP                24
+#define R_LARCH_SOP_PUSH_GPREL              25
+#define R_LARCH_SOP_PUSH_TLS_TPREL          26
+#define R_LARCH_SOP_PUSH_TLS_GOT            27
+#define R_LARCH_SOP_PUSH_TLS_GD             28
+#define R_LARCH_SOP_PUSH_PLT_PCREL          29
+#define R_LARCH_SOP_ASSERT                  30
+#define R_LARCH_SOP_NOT                     31
+#define R_LARCH_SOP_SUB                     32
+#define R_LARCH_SOP_SL                      33
+#define R_LARCH_SOP_SR                      34
+#define R_LARCH_SOP_ADD                     35
+#define R_LARCH_SOP_AND                     36
+#define R_LARCH_SOP_IF_ELSE                 37
+#define R_LARCH_SOP_POP_32_S_10_5           38
+#define R_LARCH_SOP_POP_32_U_10_12          39
+#define R_LARCH_SOP_POP_32_S_10_12          40
+#define R_LARCH_SOP_POP_32_S_10_16          41
+#define R_LARCH_SOP_POP_32_S_10_16_S2       42
+#define R_LARCH_SOP_POP_32_S_5_20           43
+#define R_LARCH_SOP_POP_32_S_0_5_10_16_S2   44
+#define R_LARCH_SOP_POP_32_S_0_10_10_16_S2  45
+#define R_LARCH_SOP_POP_32_U                46
+#define R_LARCH_ADD8                        47
+#define R_LARCH_ADD16                       48
+#define R_LARCH_ADD24                       49
+#define R_LARCH_ADD32                       50
+#define R_LARCH_ADD64                       51
+#define R_LARCH_SUB8                        52
+#define R_LARCH_SUB16                       53
+#define R_LARCH_SUB24                       54
+#define R_LARCH_SUB32                       55
+#define R_LARCH_SUB64                       56
+#define R_LARCH_GNU_VTINHERIT               57
+#define R_LARCH_GNU_VTENTRY                 58
+#define R_LARCH_B16                         64
+#define R_LARCH_B21                         65
+#define R_LARCH_B26                         66
+#define R_LARCH_ABS_HI20                    67
+#define R_LARCH_ABS_LO12                    68
+#define R_LARCH_ABS64_LO20                  69
+#define R_LARCH_ABS64_HI12                  70
+#define R_LARCH_PCALA_HI20                  71
+#define R_LARCH_PCALA_LO12                  72
+#define R_LARCH_PCALA64_LO20                73
+#define R_LARCH_PCALA64_HI12                74
+#define R_LARCH_GOT_PC_HI20                 75
+#define R_LARCH_GOT_PC_LO12                 76
+#define R_LARCH_GOT64_PC_LO20               77
+#define R_LARCH_GOT64_PC_HI12               78
+#define R_LARCH_GOT_HI20                    79
+#define R_LARCH_GOT_LO12                    80
+#define R_LARCH_GOT64_LO20                  81
+#define R_LARCH_GOT64_HI12                  82
+#define R_LARCH_TLS_LE_HI20                 83
+#define R_LARCH_TLS_LE_LO12                 84
+#define R_LARCH_TLS_LE64_LO20               85
+#define R_LARCH_TLS_LE64_HI12               86
+#define R_LARCH_TLS_IE_PC_HI20              87
+#define R_LARCH_TLS_IE_PC_LO12              88
+#define R_LARCH_TLS_IE64_PC_LO20            89
+#define R_LARCH_TLS_IE64_PC_HI12            90
+#define R_LARCH_TLS_IE_HI20                 91
+#define R_LARCH_TLS_IE_LO12                 92
+#define R_LARCH_TLS_IE64_LO20               93
+#define R_LARCH_TLS_IE64_HI12               94
+#define R_LARCH_TLS_LD_PC_HI20              95
+#define R_LARCH_TLS_LD_HI20                 96
+#define R_LARCH_TLS_GD_PC_HI20              97
+#define R_LARCH_TLS_GD_HI20                 98
+#define R_LARCH_32_PCREL                    99
+#define R_LARCH_RELAX                       100
+
 #ifdef __cplusplus
 }
 #endif
diff --git a/src/fenv/loongarch64/fenv.S b/src/fenv/loongarch64/fenv.S
new file mode 100644
index 00000000..54064e01
--- /dev/null
+++ b/src/fenv/loongarch64/fenv.S
@@ -0,0 +1,78 @@
+#ifndef __loongarch_soft_float
+
+#ifdef BROKEN_LOONGARCH_FCSR_ASM
+#define FCSR $r0
+#else
+#define FCSR $fcsr0
+#endif
+
+.global	feclearexcept
+.type   feclearexcept,@function
+feclearexcept:
+	li.w    $t0, 0x1f0000
+	and     $a0, $a0, $t0
+	movfcsr2gr $t1, FCSR
+	andn    $t1, $t1, $a0
+	movgr2fcsr FCSR, $t1
+	li.w    $a0, 0
+	jr      $ra
+
+.global	feraiseexcept
+.type   feraiseexcept,@function
+feraiseexcept:
+	li.w    $t0, 0x1f0000
+	and     $a0, $a0, $t0
+	movfcsr2gr $t1, FCSR
+	or      $t1, $t1, $a0
+	movgr2fcsr FCSR, $t1
+	li.w    $a0, 0
+	jr      $ra
+
+.global	fetestexcept
+.type   fetestexcept,@function
+fetestexcept:
+	li.w    $t0, 0x1f0000
+	and     $a0, $a0, $t0
+	movfcsr2gr $t1, FCSR
+	and     $a0, $t1, $a0
+	jr      $ra
+
+.global	fegetround
+.type   fegetround,@function
+fegetround:
+	movfcsr2gr $t0, FCSR
+	andi    $a0, $t0, 0x300
+	jr      $ra
+
+.global	__fesetround
+.hidden __fesetround
+.type   __fesetround,@function
+__fesetround:
+	li.w    $t0, 0x300
+	and     $a0, $a0, $t0
+	movfcsr2gr $t1, FCSR
+	andn    $t1, $t1, $t0
+	or      $t1, $t1, $a0
+	movgr2fcsr FCSR, $t1
+	li.w    $a0, 0
+	jr      $ra
+
+.global	fegetenv
+.type   fegetenv,@function
+fegetenv:
+	movfcsr2gr $t0, FCSR
+	st.w    $t0, $a0, 0
+	li.w    $a0, 0
+	jr      $ra
+
+.global	fesetenv
+.type   fesetenv,@function
+fesetenv:
+	addi.d  $t0, $a0, 1
+	beq     $t0, $r0, 1f
+	ld.w    $t0, $a0, 0
+1:	movgr2fcsr FCSR, $t0
+	li.w    $a0, 0
+	jr      $ra
+
+#endif
diff --git a/src/ldso/loongarch64/dlsym.s b/src/ldso/loongarch64/dlsym.s
new file mode 100644
index 00000000..26fabcdb
--- /dev/null
+++ b/src/ldso/loongarch64/dlsym.s
@@ -0,0 +1,7 @@
+.global dlsym
+.hidden __dlsym
+.type   dlsym,@function
+dlsym:
+	move      $a2, $ra
+	la.global $t0, __dlsym
+	jr        $t0
diff --git a/src/setjmp/loongarch64/longjmp.S b/src/setjmp/loongarch64/longjmp.S
new file mode 100644
index 00000000..896d2e26
--- /dev/null
+++ b/src/setjmp/loongarch64/longjmp.S
@@ -0,0 +1,32 @@
+.global _longjmp
+.global longjmp
+.type   _longjmp,@function
+.type   longjmp,@function
+_longjmp:
+longjmp:
+	ld.d    $ra, $a0, 0
+	ld.d    $sp, $a0, 8
+	ld.d    $r21,$a0, 16
+	ld.d    $fp, $a0, 24
+	ld.d    $s0, $a0, 32
+	ld.d    $s1, $a0, 40
+	ld.d    $s2, $a0, 48
+	ld.d    $s3, $a0, 56
+	ld.d    $s4, $a0, 64
+	ld.d    $s5, $a0, 72
+	ld.d    $s6, $a0, 80
+	ld.d    $s7, $a0, 88
+	ld.d    $s8, $a0, 96
+#ifndef __loongarch_soft_float
+	fld.d   $fs0, $a0, 104
+	fld.d   $fs1, $a0, 112
+	fld.d   $fs2, $a0, 120
+	fld.d   $fs3, $a0, 128
+	fld.d   $fs4, $a0, 136
+	fld.d   $fs5, $a0, 144
+	fld.d   $fs6, $a0, 152
+	fld.d   $fs7, $a0, 160
+#endif
+	sltui   $a0, $a1, 1
+	add.d   $a0, $a0, $a1
+	jr      $ra
diff --git a/src/setjmp/loongarch64/setjmp.S b/src/setjmp/loongarch64/setjmp.S
new file mode 100644
index 00000000..d158a3d2
--- /dev/null
+++ b/src/setjmp/loongarch64/setjmp.S
@@ -0,0 +1,34 @@
+.global __setjmp
+.global _setjmp
+.global setjmp
+.type __setjmp,@function
+.type _setjmp,@function
+.type setjmp,@function
+__setjmp:
+_setjmp:
+setjmp:
+	st.d    $ra, $a0, 0
+	st.d    $sp, $a0, 8
+	st.d    $r21,$a0, 16
+	st.d    $fp, $a0, 24
+	st.d    $s0, $a0, 32
+	st.d    $s1, $a0, 40
+	st.d    $s2, $a0, 48
+	st.d    $s3, $a0, 56
+	st.d    $s4, $a0, 64
+	st.d    $s5, $a0, 72
+	st.d    $s6, $a0, 80
+	st.d    $s7, $a0, 88
+	st.d    $s8, $a0, 96
+#ifndef __loongarch_soft_float
+	fst.d   $fs0, $a0, 104
+	fst.d   $fs1, $a0, 112
+	fst.d   $fs2, $a0, 120
+	fst.d   $fs3, $a0, 128
+	fst.d   $fs4, $a0, 136
+	fst.d   $fs5, $a0, 144
+	fst.d   $fs6, $a0, 152
+	fst.d   $fs7, $a0, 160
+#endif
+	move    $a0, $zero
+	jr      $ra
diff --git a/src/signal/loongarch64/restore.s b/src/signal/loongarch64/restore.s
new file mode 100644
index 00000000..f8e6daeb
--- /dev/null
+++ b/src/signal/loongarch64/restore.s
@@ -0,0 +1,10 @@
+.global	__restore_rt
+.global	__restore
+.hidden __restore_rt
+.hidden __restore
+.type   __restore_rt,@function
+.type   __restore,@function
+__restore_rt:
+__restore:
+	li.w    $a7, 139
+	syscall	0
diff --git a/src/signal/loongarch64/sigsetjmp.s b/src/signal/loongarch64/sigsetjmp.s
new file mode 100644
index 00000000..992ab1a4
--- /dev/null
+++ b/src/signal/loongarch64/sigsetjmp.s
@@ -0,0 +1,25 @@
+.global sigsetjmp
+.global __sigsetjmp
+.type sigsetjmp,@function
+.type __sigsetjmp,@function
+sigsetjmp:
+__sigsetjmp:
+	beq     $a1, $zero, 1f
+	st.d    $ra, $a0, 168
+	st.d    $s0, $a0, 176
+	move    $s0, $a0
+
+	la.global  $t0, setjmp
+	jirl       $ra, $t0, 0
+
+	move    $a1, $a0        # Return from 'setjmp' or 'longjmp'
+	move    $a0, $s0
+	ld.d    $ra, $a0, 168
+	ld.d    $s0, $a0, 176
+
+.hidden __sigsetjmp_tail
+	la.global  $t0, __sigsetjmp_tail
+	jr         $t0
+1:
+	la.global  $t0, setjmp
+	jr         $t0
diff --git a/src/thread/loongarch64/__set_thread_area.s b/src/thread/loongarch64/__set_thread_area.s
new file mode 100644
index 00000000..ffdd52f1
--- /dev/null
+++ b/src/thread/loongarch64/__set_thread_area.s
@@ -0,0 +1,7 @@
+.global	__set_thread_area
+.hidden __set_thread_area
+.type   __set_thread_area,@function
+__set_thread_area:
+	move $tp, $a0
+	move $a0, $zero
+	jr   $ra
diff --git a/src/thread/loongarch64/__unmapself.s b/src/thread/loongarch64/__unmapself.s
new file mode 100644
index 00000000..1de334af
--- /dev/null
+++ b/src/thread/loongarch64/__unmapself.s
@@ -0,0 +1,7 @@
+.global	__unmapself
+.type   __unmapself, @function
+__unmapself:
+	li.d    $a7, 215   # call munmap
+	syscall	0
+	li.d    $a7, 93    # call exit
+	syscall	0
diff --git a/src/thread/loongarch64/clone.s b/src/thread/loongarch64/clone.s
new file mode 100644
index 00000000..db9015e6
--- /dev/null
+++ b/src/thread/loongarch64/clone.s
@@ -0,0 +1,28 @@
+#__clone(func, stack, flags, arg, ptid, tls, ctid)
+#         a0,    a1,   a2,    a3,  a4,  a5,   a6
+# sys_clone(flags, stack, ptid, ctid, tls)
+#            a0,    a1,   a2,    a3,  a4
+
+.global	__clone
+.hidden __clone
+.type __clone,@function
+__clone:
+	# Save function pointer and argument pointer on new thread stack
+	addi.d  $a1, $a1, -16
+	st.d    $a0, $a1, 0     # save function pointer
+	st.d    $a3, $a1, 8     # save argument pointer
+	or      $a0, $a2, $zero
+	or      $a2, $a4, $zero
+	or      $a3, $a6, $zero
+	or      $a4, $a5, $zero
+	ori     $a7, $zero, 220
+	syscall 0               # call clone
+
+	beqz    $a0, 1f         # whether child process
+	jirl    $zero, $ra, 0   # parent process return
+1:
+	ld.d    $t8, $sp, 0     # function pointer
+	ld.d    $a0, $sp, 8     # argument pointer
+	jirl    $ra, $t8, 0     # call the user's function
+	ori     $a7, $zero, 93
+	syscall	0               # child process exit
diff --git a/src/thread/loongarch64/syscall_cp.s b/src/thread/loongarch64/syscall_cp.s
new file mode 100644
index 00000000..0fbc7a47
--- /dev/null
+++ b/src/thread/loongarch64/syscall_cp.s
@@ -0,0 +1,29 @@
+.global	__cp_begin
+.hidden	__cp_begin
+.global	__cp_end
+.hidden	__cp_end
+.global	__cp_cancel
+.hidden	__cp_cancel
+.hidden	__cancel
+.global	__syscall_cp_asm
+.hidden	__syscall_cp_asm
+.type   __syscall_cp_asm,@function
+
+__syscall_cp_asm:
+__cp_begin:
+	ld.w $a0, $a0, 0
+	bnez $a0, __cp_cancel
+	move $t8, $a1    # reserve system call number
+	move $a0, $a2
+	move $a1, $a3
+	move $a2, $a4
+	move $a3, $a5
+	move $a4, $a6
+	move $a5, $a7
+	move $a7, $t8
+	syscall	0
+__cp_end:
+	jr $ra
+__cp_cancel:
+	la.local $t8, __cancel
+	jr $t8

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [musl] loongarch64 merge
  2024-01-25 17:43 [musl] loongarch64 merge Rich Felker
@ 2024-01-26  1:18 ` Fangrui Song
  2024-01-26  1:37   ` Rich Felker
  2024-01-29 16:47 ` Rich Felker
  1 sibling, 1 reply; 7+ messages in thread
From: Fangrui Song @ 2024-01-26  1:18 UTC (permalink / raw)
  To: musl

On Thu, Jan 25, 2024 at 9:43 AM Rich Felker <dalias@libc.org> wrote:
>
> I'm going through where everything was left on this topic and
> preparing a patch for merge. This message/thread is to document what
> I'm actually doing vs the various submitted versions of the patch
> since v5/v6 where the major review took place.
>
>
> Subsequent changes I'm reverting:
>
> - De-optimization of __get_tp. No motivation for removing the
>   potentially in-place $tp was provided, and we generally use the
>   arch's tp in-place unless there's a compiler bug to be worked
>   around. See powerpc{,64} for an example where it's used, or1k where
>   we have a probably-obsolete workaround for ancient clang being
>   broken.
>
> - unsigned -> unsigned int, etc.
>
> - Gratuitous whitespace changes in headers that obscure the fact that
>   a header is a complete duplicate that could eventually be shared
>   between archs (e.g. bits/float.h, bits/posix.h) or just obscure
>   what differs from other archs when running diff.
>
>
> Fixes from previous review that were overlooked:
>
> - Removing SA_RESTORER -- its presence defined as 0 produces wrong
>   sigaction ABI.
>
>
> Additions:
>
> - Adding the reloc.h/configure case for single-only float.
>
> - The new member names for mcontext_t are all in reserved namespace,
>   so there's no reason to have a separate namespace-clean version of
>   mcontext_t, and I'm removing the latter.
>
> - Public member uc_flags with no __, macro for compat with any
>   existing software using the __-prefixed name.
>
>
> Still TODO:
>
> I don't think I ever reviewed the apparent rewrite of sigsetjmp and
> possibly some other asm that changed between v5 and v8. I'm about to
> start looking at that and will follow up.
>
>
> Attached are a "differences vs v8" patch and what my cumulative patch
> looks like right now.
>
> Rich

There are a few new relocation types, including R_LARCH_ALIGN (102),
some ULEB128 ones, and some TLSDESC ones.

https://github.com/loongson/la-abi-specs/blob/release/laelf.adoc

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [musl] loongarch64 merge
  2024-01-26  1:18 ` Fangrui Song
@ 2024-01-26  1:37   ` Rich Felker
  0 siblings, 0 replies; 7+ messages in thread
From: Rich Felker @ 2024-01-26  1:37 UTC (permalink / raw)
  To: Fangrui Song; +Cc: musl

On Thu, Jan 25, 2024 at 05:18:21PM -0800, Fangrui Song wrote:
> On Thu, Jan 25, 2024 at 9:43 AM Rich Felker <dalias@libc.org> wrote:
> >
> > I'm going through where everything was left on this topic and
> > preparing a patch for merge. This message/thread is to document what
> > I'm actually doing vs the various submitted versions of the patch
> > since v5/v6 where the major review took place.
> >
> >
> > Subsequent changes I'm reverting:
> >
> > - De-optimization of __get_tp. No motivation for removing the
> >   potentially in-place $tp was provided, and we generally use the
> >   arch's tp in-place unless there's a compiler bug to be worked
> >   around. See powerpc{,64} for an example where it's used, or1k where
> >   we have a probably-obsolete workaround for ancient clang being
> >   broken.
> >
> > - unsigned -> unsigned int, etc.
> >
> > - Gratuitous whitespace changes in headers that obscure the fact that
> >   a header is a complete duplicate that could eventually be shared
> >   between archs (e.g. bits/float.h, bits/posix.h) or just obscure
> >   what differs from other archs when running diff.
> >
> >
> > Fixes from previous review that were overlooked:
> >
> > - Removing SA_RESTORER -- its presence defined as 0 produces wrong
> >   sigaction ABI.
> >
> >
> > Additions:
> >
> > - Adding the reloc.h/configure case for single-only float.
> >
> > - The new member names for mcontext_t are all in reserved namespace,
> >   so there's no reason to have a separate namespace-clean version of
> >   mcontext_t, and I'm removing the latter.
> >
> > - Public member uc_flags with no __, macro for compat with any
> >   existing software using the __-prefixed name.
> >
> >
> > Still TODO:
> >
> > I don't think I ever reviewed the apparent rewrite of sigsetjmp and
> > possibly some other asm that changed between v5 and v8. I'm about to
> > start looking at that and will follow up.
> >
> >
> > Attached are a "differences vs v8" patch and what my cumulative patch
> > looks like right now.
> >
> > Rich
> 
> There are a few new relocation types, including R_LARCH_ALIGN (102),
> some ULEB128 ones, and some TLSDESC ones.
> 
> https://github.com/loongson/la-abi-specs/blob/release/laelf.adoc

I'd kinda wished the elf.h stuff had been done as a separate patch
prepping for the port (since having new archs' elf constants in elf.h
has value even without a musl port to the arch). If you (or anyone
else) want to propose patches for this or any other new archs we're
missing, I'd be happy to separate that out, but unless it's ready
right away I don't want to hold up this port any longer on minor
details. I mainly just want to make sure what gets merged (1) doesn't
have broken ABI, (2) follows policy, and (3) is reasonably easy to
analyze to see what actually differs between archs, and easy to follow
any further change history made later. Getting back to this as I'd
promised was really the last big thing holding up a release, which I
know folks have been eagar for for a long time, and I'd like to move
on with it asap and not let this get put off again.

Rich

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [musl] loongarch64 merge
  2024-01-25 17:43 [musl] loongarch64 merge Rich Felker
  2024-01-26  1:18 ` Fangrui Song
@ 2024-01-29 16:47 ` Rich Felker
  2024-01-30  7:35   ` Hongliang Wang
  1 sibling, 1 reply; 7+ messages in thread
From: Rich Felker @ 2024-01-29 16:47 UTC (permalink / raw)
  To: musl; +Cc: Hongliang Wang

On Thu, Jan 25, 2024 at 12:43:54PM -0500, Rich Felker wrote:
> I'm going through where everything was left on this topic and
> preparing a patch for merge. This message/thread is to document what
> I'm actually doing vs the various submitted versions of the patch
> since v5/v6 where the major review took place.
> 
> 
> Subsequent changes I'm reverting:
> 
> - De-optimization of __get_tp. No motivation for removing the
>   potentially in-place $tp was provided, and we generally use the
>   arch's tp in-place unless there's a compiler bug to be worked
>   around. See powerpc{,64} for an example where it's used, or1k where
>   we have a probably-obsolete workaround for ancient clang being
>   broken.
> 
> - unsigned -> unsigned int, etc.
> 
> - Gratuitous whitespace changes in headers that obscure the fact that
>   a header is a complete duplicate that could eventually be shared
>   between archs (e.g. bits/float.h, bits/posix.h) or just obscure
>   what differs from other archs when running diff.
> 
> 
> Fixes from previous review that were overlooked:
> 
> - Removing SA_RESTORER -- its presence defined as 0 produces wrong
>   sigaction ABI.
> 
> 
> Additions:
> 
> - Adding the reloc.h/configure case for single-only float.
> 
> - The new member names for mcontext_t are all in reserved namespace,
>   so there's no reason to have a separate namespace-clean version of
>   mcontext_t, and I'm removing the latter.
> 
> - Public member uc_flags with no __, macro for compat with any
>   existing software using the __-prefixed name.
> 
> 
> Still TODO:
> 
> I don't think I ever reviewed the apparent rewrite of sigsetjmp and
> possibly some other asm that changed between v5 and v8. I'm about to
> start looking at that and will follow up.

OK, sigsetjmp appears to have been non-working in v5 (it failed to
save the argument anywhere, so attempted it use it after it was
clobbered in the code path after calling setjmp).

v8 fixes this and seems like it should work, but I don't understand
the offsets used.

The loongarch64 __jmp_buf has 23 slots, but only appears to use 21 of
them, with sigsetjmp storing its extra data in the last 2 unused
slots. The contract here is supposed to be that the entirety of
__jmp_buf belongs to setjmp/longjmp, with sigsetjmp using the extra
space in the full jmp_buf/sigjmp_buf type with signal mask and lots of
extra space.

Were these slots added just for sigsetjmp? If so, that was an ABI
mistake, but I don't think it warrants a late change; it's probably
better to just leave them there as extensibility.

Either way, I think sigsetjmp should be modified not to rely on
storage in space that setjmp is entitled to be able to modify, i.e. it
should be using offsets 184 and 184+16, not 168 and 176.

Rich

^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [musl] loongarch64 merge
  2024-01-29 16:47 ` Rich Felker
@ 2024-01-30  7:35   ` Hongliang Wang
  2024-02-23  3:26     ` Hongliang Wang
  0 siblings, 1 reply; 7+ messages in thread
From: Hongliang Wang @ 2024-01-30  7:35 UTC (permalink / raw)
  To: Rich Felker, musl

[-- Attachment #1: Type: text/plain, Size: 3949 bytes --]



在 2024/1/30 上午12:47, Rich Felker 写道:
> On Thu, Jan 25, 2024 at 12:43:54PM -0500, Rich Felker wrote:
>> I'm going through where everything was left on this topic and
>> preparing a patch for merge. This message/thread is to document what
>> I'm actually doing vs the various submitted versions of the patch
>> since v5/v6 where the major review took place.
>>
>>
>> Subsequent changes I'm reverting:
>>
>> - De-optimization of __get_tp. No motivation for removing the
>>    potentially in-place $tp was provided, and we generally use the
>>    arch's tp in-place unless there's a compiler bug to be worked
>>    around. See powerpc{,64} for an example where it's used, or1k where
>>    we have a probably-obsolete workaround for ancient clang being
>>    broken.
>>
>> - unsigned -> unsigned int, etc.
>>
>> - Gratuitous whitespace changes in headers that obscure the fact that
>>    a header is a complete duplicate that could eventually be shared
>>    between archs (e.g. bits/float.h, bits/posix.h) or just obscure
>>    what differs from other archs when running diff.
>>
>>
>> Fixes from previous review that were overlooked:
>>
>> - Removing SA_RESTORER -- its presence defined as 0 produces wrong
>>    sigaction ABI.
>>
>>
>> Additions:
>>
>> - Adding the reloc.h/configure case for single-only float.
>>
>> - The new member names for mcontext_t are all in reserved namespace,
>>    so there's no reason to have a separate namespace-clean version of
>>    mcontext_t, and I'm removing the latter.
>>
>> - Public member uc_flags with no __, macro for compat with any
>>    existing software using the __-prefixed name.
>>
>>
>> Still TODO:
>>
>> I don't think I ever reviewed the apparent rewrite of sigsetjmp and
>> possibly some other asm that changed between v5 and v8. I'm about to
>> start looking at that and will follow up.
> 
> OK, sigsetjmp appears to have been non-working in v5 (it failed to
> save the argument anywhere, so attempted it use it after it was
> clobbered in the code path after calling setjmp).
> 
> v8 fixes this and seems like it should work, but I don't understand
> the offsets used.
> 
> The loongarch64 __jmp_buf has 23 slots, but only appears to use 21 of
> them, with sigsetjmp storing its extra data in the last 2 unused
> slots. The contract here is supposed to be that the entirety of
> __jmp_buf belongs to setjmp/longjmp, with sigsetjmp using the extra
> space in the full jmp_buf/sigjmp_buf type with signal mask and lots of
> extra space.
> 
> Were these slots added just for sigsetjmp? If so, that was an ABI
> mistake, but I don't think it warrants a late change; it's probably
> better to just leave them there as extensibility.

Yes, the purpose of the last 2 slots of __jmp_buf just used for
sigsetjmp to storing its extra data.
Now leave them there as extensibility is OK.

Hongliang Wang

> 
> Either way, I think sigsetjmp should be modified not to rely on
> storage in space that setjmp is entitled to be able to modify, i.e. it
> should be using offsets 184 and 184+16, not 168 and 176.
> 
> Rich
> 
I understand. I attached a patch to fix this issue. please review it.

diff --git a/src/signal/loongarch64/sigsetjmp.s 
b/src/signal/loongarch64/sigsetjmp.s
index 992ab1a4..9c0e3ae2 100644
--- a/src/signal/loongarch64/sigsetjmp.s
+++ b/src/signal/loongarch64/sigsetjmp.s
@@ -5,8 +5,8 @@
  sigsetjmp:
  __sigsetjmp:
         beq     $a1, $zero, 1f
-       st.d    $ra, $a0, 168
-       st.d    $s0, $a0, 176
+       st.d    $ra, $a0, 184
+       st.d    $s0, $a0, 200  #184+8+8
         move    $s0, $a0

         la.global  $t0, setjmp
@@ -14,8 +14,8 @@ __sigsetjmp:

         move    $a1, $a0        # Return from 'setjmp' or 'longjmp'
         move    $a0, $s0
-       ld.d    $ra, $a0, 168
-       ld.d    $s0, $a0, 176
+       ld.d    $ra, $a0, 184
+       ld.d    $s0, $a0, 200 #184+8+8

  .hidden __sigsetjmp_tail
         la.global  $t0, __sigsetjmp_tail
(END)

Hongliang Wang

[-- Attachment #2: 0001-loongarch64-fix-the-incorrect-offset-of__jmp_buf-use.patch --]
[-- Type: text/x-patch, Size: 1116 bytes --]

From 762a21eb0e1addf508cdb3ab2e3c109400734452 Mon Sep 17 00:00:00 2001
From: wanghongliang <wanghongliang@loongson.cn>
Date: Tue, 30 Jan 2024 06:51:56 +0800
Subject: [PATCH] loongarch64:fix the incorrect offset of__jmp_buf used in
 sigsetjmp.

Signed-off-by: wanghongliang <wanghongliang@loongson.cn>
---
 src/signal/loongarch64/sigsetjmp.s | 8 ++++----
 1 file changed, 4 insertions(+), 4 deletions(-)

diff --git a/src/signal/loongarch64/sigsetjmp.s b/src/signal/loongarch64/sigsetjmp.s
index 992ab1a4..9c0e3ae2 100644
--- a/src/signal/loongarch64/sigsetjmp.s
+++ b/src/signal/loongarch64/sigsetjmp.s
@@ -5,8 +5,8 @@
 sigsetjmp:
 __sigsetjmp:
 	beq     $a1, $zero, 1f
-	st.d    $ra, $a0, 168
-	st.d    $s0, $a0, 176
+	st.d    $ra, $a0, 184
+	st.d    $s0, $a0, 200  #184+8+8
 	move    $s0, $a0
 
 	la.global  $t0, setjmp
@@ -14,8 +14,8 @@ __sigsetjmp:
 
 	move    $a1, $a0        # Return from 'setjmp' or 'longjmp'
 	move    $a0, $s0
-	ld.d    $ra, $a0, 168
-	ld.d    $s0, $a0, 176
+	ld.d    $ra, $a0, 184
+	ld.d    $s0, $a0, 200 #184+8+8
 
 .hidden __sigsetjmp_tail
 	la.global  $t0, __sigsetjmp_tail
-- 
2.37.1


^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [musl] loongarch64 merge
  2024-01-30  7:35   ` Hongliang Wang
@ 2024-02-23  3:26     ` Hongliang Wang
  2024-02-23 15:36       ` Rich Felker
  0 siblings, 1 reply; 7+ messages in thread
From: Hongliang Wang @ 2024-02-23  3:26 UTC (permalink / raw)
  To: Rich Felker, musl

Hi, Rich

We are very excited to see that LoongArch has been merged in musl.
Next, we will continue to work hard to improve LoongArch, and we
still require your strong support.

Thank you for your help on LoongArch.

Regards,
Hongliang Wang

在 2024/1/30 下午3:35, Hongliang Wang 写道:
> 
> 
> 在 2024/1/30 上午12:47, Rich Felker 写道:
>> On Thu, Jan 25, 2024 at 12:43:54PM -0500, Rich Felker wrote:
>>> I'm going through where everything was left on this topic and
>>> preparing a patch for merge. This message/thread is to document what
>>> I'm actually doing vs the various submitted versions of the patch
>>> since v5/v6 where the major review took place.
>>>
>>>
>>> Subsequent changes I'm reverting:
>>>
>>> - De-optimization of __get_tp. No motivation for removing the
>>>    potentially in-place $tp was provided, and we generally use the
>>>    arch's tp in-place unless there's a compiler bug to be worked
>>>    around. See powerpc{,64} for an example where it's used, or1k where
>>>    we have a probably-obsolete workaround for ancient clang being
>>>    broken.
>>>
>>> - unsigned -> unsigned int, etc.
>>>
>>> - Gratuitous whitespace changes in headers that obscure the fact that
>>>    a header is a complete duplicate that could eventually be shared
>>>    between archs (e.g. bits/float.h, bits/posix.h) or just obscure
>>>    what differs from other archs when running diff.
>>>
>>>
>>> Fixes from previous review that were overlooked:
>>>
>>> - Removing SA_RESTORER -- its presence defined as 0 produces wrong
>>>    sigaction ABI.
>>>
>>>
>>> Additions:
>>>
>>> - Adding the reloc.h/configure case for single-only float.
>>>
>>> - The new member names for mcontext_t are all in reserved namespace,
>>>    so there's no reason to have a separate namespace-clean version of
>>>    mcontext_t, and I'm removing the latter.
>>>
>>> - Public member uc_flags with no __, macro for compat with any
>>>    existing software using the __-prefixed name.
>>>
>>>
>>> Still TODO:
>>>
>>> I don't think I ever reviewed the apparent rewrite of sigsetjmp and
>>> possibly some other asm that changed between v5 and v8. I'm about to
>>> start looking at that and will follow up.
>>
>> OK, sigsetjmp appears to have been non-working in v5 (it failed to
>> save the argument anywhere, so attempted it use it after it was
>> clobbered in the code path after calling setjmp).
>>
>> v8 fixes this and seems like it should work, but I don't understand
>> the offsets used.
>>
>> The loongarch64 __jmp_buf has 23 slots, but only appears to use 21 of
>> them, with sigsetjmp storing its extra data in the last 2 unused
>> slots. The contract here is supposed to be that the entirety of
>> __jmp_buf belongs to setjmp/longjmp, with sigsetjmp using the extra
>> space in the full jmp_buf/sigjmp_buf type with signal mask and lots of
>> extra space.
>>
>> Were these slots added just for sigsetjmp? If so, that was an ABI
>> mistake, but I don't think it warrants a late change; it's probably
>> better to just leave them there as extensibility.
> 
> Yes, the purpose of the last 2 slots of __jmp_buf just used for
> sigsetjmp to storing its extra data.
> Now leave them there as extensibility is OK.
> 
> Hongliang Wang
> 
>>
>> Either way, I think sigsetjmp should be modified not to rely on
>> storage in space that setjmp is entitled to be able to modify, i.e. it
>> should be using offsets 184 and 184+16, not 168 and 176.
>>
>> Rich
>>
> I understand. I attached a patch to fix this issue. please review it.
> 
> diff --git a/src/signal/loongarch64/sigsetjmp.s 
> b/src/signal/loongarch64/sigsetjmp.s
> index 992ab1a4..9c0e3ae2 100644
> --- a/src/signal/loongarch64/sigsetjmp.s
> +++ b/src/signal/loongarch64/sigsetjmp.s
> @@ -5,8 +5,8 @@
>   sigsetjmp:
>   __sigsetjmp:
>          beq     $a1, $zero, 1f
> -       st.d    $ra, $a0, 168
> -       st.d    $s0, $a0, 176
> +       st.d    $ra, $a0, 184
> +       st.d    $s0, $a0, 200  #184+8+8
>          move    $s0, $a0
> 
>          la.global  $t0, setjmp
> @@ -14,8 +14,8 @@ __sigsetjmp:
> 
>          move    $a1, $a0        # Return from 'setjmp' or 'longjmp'
>          move    $a0, $s0
> -       ld.d    $ra, $a0, 168
> -       ld.d    $s0, $a0, 176
> +       ld.d    $ra, $a0, 184
> +       ld.d    $s0, $a0, 200 #184+8+8
> 
>   .hidden __sigsetjmp_tail
>          la.global  $t0, __sigsetjmp_tail
> (END)
> 
> Hongliang Wang


^ permalink raw reply	[flat|nested] 7+ messages in thread

* Re: [musl] loongarch64 merge
  2024-02-23  3:26     ` Hongliang Wang
@ 2024-02-23 15:36       ` Rich Felker
  0 siblings, 0 replies; 7+ messages in thread
From: Rich Felker @ 2024-02-23 15:36 UTC (permalink / raw)
  To: Hongliang Wang; +Cc: musl

On Fri, Feb 23, 2024 at 11:26:38AM +0800, Hongliang Wang wrote:
> Hi, Rich
> 
> We are very excited to see that LoongArch has been merged in musl.
> Next, we will continue to work hard to improve LoongArch, and we
> still require your strong support.
> 
> Thank you for your help on LoongArch.

Yes, I'm glad to finally have it done -- I didn't want to let another
release go by without it, and I'm trying to put together the release.
Hopefully we can get some independent smoke testing done between now
and when the relase happens.

Rich

^ permalink raw reply	[flat|nested] 7+ messages in thread

end of thread, other threads:[~2024-02-23 15:36 UTC | newest]

Thread overview: 7+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2024-01-25 17:43 [musl] loongarch64 merge Rich Felker
2024-01-26  1:18 ` Fangrui Song
2024-01-26  1:37   ` Rich Felker
2024-01-29 16:47 ` Rich Felker
2024-01-30  7:35   ` Hongliang Wang
2024-02-23  3:26     ` Hongliang Wang
2024-02-23 15:36       ` Rich Felker

Code repositories for project(s) associated with this public inbox

	https://git.vuxu.org/mirror/musl/

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 NNTP newsgroup(s).