glibc/sysdeps/unix/sysv/linux/arm/sysdep.h
Paul Eggert 581c785bf3 Update copyright dates with scripts/update-copyrights
I used these shell commands:

../glibc/scripts/update-copyrights $PWD/../gnulib/build-aux/update-copyright
(cd ../glibc && git commit -am"[this commit message]")

and then ignored the output, which consisted lines saying "FOO: warning:
copyright statement not found" for each of 7061 files FOO.

I then removed trailing white space from math/tgmath.h,
support/tst-support-open-dev-null-range.c, and
sysdeps/x86_64/multiarch/strlen-vec.S, to work around the following
obscure pre-commit check failure diagnostics from Savannah.  I don't
know why I run into these diagnostics whereas others evidently do not.

remote: *** 912-#endif
remote: *** 913:
remote: *** 914-
remote: *** error: lines with trailing whitespace found
...
remote: *** error: sysdeps/unix/sysv/linux/statx_cp.c: trailing lines
2022-01-01 11:40:24 -08:00

416 lines
12 KiB
C

/* Copyright (C) 1992-2022 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_ARM_SYSDEP_H
#define _LINUX_ARM_SYSDEP_H 1
/* There is some commonality. */
#include <sysdeps/unix/sysv/linux/sysdep.h>
#include <sysdeps/unix/arm/sysdep.h>
/* Defines RTLD_PRIVATE_ERRNO and USE_DL_SYSINFO. */
#include <dl-sysdep.h>
#include <tls.h>
/* For Linux we can use the system call table in the header file
/usr/include/asm/unistd.h
of the kernel. But these symbols do not follow the SYS_* syntax
so we have to redefine the `SYS_ify' macro here. */
#undef SYS_ify
#define SYS_ify(syscall_name) (__NR_##syscall_name)
#include <bits/hwcap.h>
#ifdef __ASSEMBLER__
#ifndef ARCH_HAS_HARD_TP
/* Internal macro calling the linux kernel kuser_get_tls helper.
Note that in thumb mode, a constant pool break is often out of range, so
we always expand the constant inline. */
# ifdef __thumb2__
# define GET_TLS_BODY \
movw r0, #0x0fe0; \
movt r0, #0xffff; \
blx r0
# else
# define GET_TLS_BODY \
mov r0, #0xffff0fff; /* Point to the high page. */ \
mov lr, pc; /* Save our return address. */ \
sub pc, r0, #31 /* Jump to the TLS entry. */
# endif
/* Helper to get the TLS base pointer. Save LR in TMP, return in R0,
and no other registers clobbered. TMP may be LR itself to indicate
that no save is necessary. */
# undef GET_TLS
# define GET_TLS(TMP) \
.ifnc TMP, lr; \
mov TMP, lr; \
cfi_register (lr, TMP); \
GET_TLS_BODY; \
mov lr, TMP; \
cfi_restore (lr); \
.else; \
GET_TLS_BODY; \
.endif
#endif /* ARCH_HAS_HARD_TP */
/* Linux uses a negative return value to indicate syscall errors,
unlike most Unices, which use the condition codes' carry flag.
Since version 2.1 the return value of a system call might be
negative even if the call succeeded. E.g., the `lseek' system call
might return a large offset. Therefore we must not anymore test
for < 0, but test for a real error by making sure the value in R0
is a real error number. Linus said he will make sure the no syscall
returns a value in -1 .. -4095 as a valid result so we can safely
test with -4095. */
#undef PSEUDO
#define PSEUDO(name, syscall_name, args) \
.text; \
ENTRY (name); \
DO_CALL (syscall_name, args); \
cmn r0, $4096;
#define PSEUDO_RET \
it cc; \
RETINSTR(cc, lr); \
b PLTJMP(SYSCALL_ERROR)
#undef ret
#define ret PSEUDO_RET
#undef PSEUDO_END
#define PSEUDO_END(name) \
SYSCALL_ERROR_HANDLER; \
END (name)
#undef PSEUDO_NOERRNO
#define PSEUDO_NOERRNO(name, syscall_name, args) \
.text; \
ENTRY (name); \
DO_CALL (syscall_name, args);
#define PSEUDO_RET_NOERRNO \
DO_RET (lr);
#undef ret_NOERRNO
#define ret_NOERRNO PSEUDO_RET_NOERRNO
#undef PSEUDO_END_NOERRNO
#define PSEUDO_END_NOERRNO(name) \
END (name)
/* The function has to return the error code. */
#undef PSEUDO_ERRVAL
#define PSEUDO_ERRVAL(name, syscall_name, args) \
.text; \
ENTRY (name) \
DO_CALL (syscall_name, args); \
rsb r0, r0, #0
#undef PSEUDO_END_ERRVAL
#define PSEUDO_END_ERRVAL(name) \
END (name)
#define ret_ERRVAL PSEUDO_RET_NOERRNO
#if !IS_IN (libc)
# define SYSCALL_ERROR __local_syscall_error
# if RTLD_PRIVATE_ERRNO
# define SYSCALL_ERROR_HANDLER \
__local_syscall_error: \
rsb r0, r0, #0; \
LDST_PCREL(str, r0, r1, C_SYMBOL_NAME(rtld_errno)); \
mvn r0, #0; \
DO_RET(lr)
# else
# if defined(__ARM_ARCH_4T__) && defined(__THUMB_INTERWORK__)
# define POP_PC \
pop { lr }; \
cfi_adjust_cfa_offset (-4); \
cfi_restore (lr); \
bx lr
# else
# define POP_PC pop { pc }
# endif
# define SYSCALL_ERROR_HANDLER \
__local_syscall_error: \
push { lr }; \
cfi_adjust_cfa_offset (4); \
cfi_rel_offset (lr, 0); \
push { r0 }; \
cfi_adjust_cfa_offset (4); \
bl PLTJMP(C_SYMBOL_NAME(__errno_location)); \
pop { r1 }; \
cfi_adjust_cfa_offset (-4); \
rsb r1, r1, #0; \
str r1, [r0]; \
mvn r0, #0; \
POP_PC;
# endif
#else
# define SYSCALL_ERROR_HANDLER /* Nothing here; code in sysdep.S is used. */
# define SYSCALL_ERROR __syscall_error
#endif
/* The ARM EABI user interface passes the syscall number in r7, instead
of in the swi. This is more efficient, because the kernel does not need
to fetch the swi from memory to find out the number; which can be painful
with separate I-cache and D-cache. Make sure to use 0 for the SWI
argument; otherwise the (optional) compatibility code for APCS binaries
may be invoked. */
/* Linux takes system call args in registers:
arg 1 r0
arg 2 r1
arg 3 r2
arg 4 r3
arg 5 r4 (this is different from the APCS convention)
arg 6 r5
arg 7 r6
The compiler is going to form a call by coming here, through PSEUDO, with
arguments
syscall number in the DO_CALL macro
arg 1 r0
arg 2 r1
arg 3 r2
arg 4 r3
arg 5 [sp]
arg 6 [sp+4]
arg 7 [sp+8]
We need to shuffle values between R4..R6 and the stack so that the
caller's v1..v3 and stack frame are not corrupted, and the kernel
sees the right arguments.
*/
/* We must save and restore r7 (call-saved) for the syscall number.
We never make function calls from inside here (only potentially
signal handlers), so we do not bother with doubleword alignment.
Just like the APCS syscall convention, the EABI syscall convention uses
r0 through r6 for up to seven syscall arguments. None are ever passed to
the kernel on the stack, although incoming arguments are on the stack for
syscalls with five or more arguments.
The assembler will convert the literal pool load to a move for most
syscalls. */
#undef DO_CALL
#define DO_CALL(syscall_name, args) \
DOARGS_##args; \
ldr r7, =SYS_ify (syscall_name); \
swi 0x0; \
UNDOARGS_##args
#undef DOARGS_0
#define DOARGS_0 \
.fnstart; \
push { r7 }; \
cfi_adjust_cfa_offset (4); \
cfi_rel_offset (r7, 0); \
.save { r7 }
#undef DOARGS_1
#define DOARGS_1 DOARGS_0
#undef DOARGS_2
#define DOARGS_2 DOARGS_0
#undef DOARGS_3
#define DOARGS_3 DOARGS_0
#undef DOARGS_4
#define DOARGS_4 DOARGS_0
#undef DOARGS_5
#define DOARGS_5 \
.fnstart; \
push {r4, r7}; \
cfi_adjust_cfa_offset (8); \
cfi_rel_offset (r4, 0); \
cfi_rel_offset (r7, 4); \
.save { r4, r7 }; \
ldr r4, [sp, #8]
#undef DOARGS_6
#define DOARGS_6 \
.fnstart; \
mov ip, sp; \
push {r4, r5, r7}; \
cfi_adjust_cfa_offset (12); \
cfi_rel_offset (r4, 0); \
cfi_rel_offset (r5, 4); \
cfi_rel_offset (r7, 8); \
.save { r4, r5, r7 }; \
ldmia ip, {r4, r5}
#undef DOARGS_7
#define DOARGS_7 \
.fnstart; \
mov ip, sp; \
push {r4, r5, r6, r7}; \
cfi_adjust_cfa_offset (16); \
cfi_rel_offset (r4, 0); \
cfi_rel_offset (r5, 4); \
cfi_rel_offset (r6, 8); \
cfi_rel_offset (r7, 12); \
.save { r4, r5, r6, r7 }; \
ldmia ip, {r4, r5, r6}
#undef UNDOARGS_0
#define UNDOARGS_0 \
pop {r7}; \
cfi_adjust_cfa_offset (-4); \
cfi_restore (r7); \
.fnend
#undef UNDOARGS_1
#define UNDOARGS_1 UNDOARGS_0
#undef UNDOARGS_2
#define UNDOARGS_2 UNDOARGS_0
#undef UNDOARGS_3
#define UNDOARGS_3 UNDOARGS_0
#undef UNDOARGS_4
#define UNDOARGS_4 UNDOARGS_0
#undef UNDOARGS_5
#define UNDOARGS_5 \
pop {r4, r7}; \
cfi_adjust_cfa_offset (-8); \
cfi_restore (r4); \
cfi_restore (r7); \
.fnend
#undef UNDOARGS_6
#define UNDOARGS_6 \
pop {r4, r5, r7}; \
cfi_adjust_cfa_offset (-12); \
cfi_restore (r4); \
cfi_restore (r5); \
cfi_restore (r7); \
.fnend
#undef UNDOARGS_7
#define UNDOARGS_7 \
pop {r4, r5, r6, r7}; \
cfi_adjust_cfa_offset (-16); \
cfi_restore (r4); \
cfi_restore (r5); \
cfi_restore (r6); \
cfi_restore (r7); \
.fnend
#else /* not __ASSEMBLER__ */
#if defined(__thumb__)
/* We can not expose the use of r7 to the compiler. GCC (as
of 4.5) uses r7 as the hard frame pointer for Thumb - although
for Thumb-2 it isn't obviously a better choice than r11.
And GCC does not support asms that conflict with the frame
pointer.
This would be easier if syscall numbers never exceeded 255,
but they do. For the moment the LOAD_ARGS_7 is sacrificed.
We can't use push/pop inside the asm because that breaks
unwinding (i.e. thread cancellation) for this frame. We can't
locally save and restore r7, because we do not know if this
function uses r7 or if it is our caller's r7; if it is our caller's,
then unwinding will fail higher up the stack. So we move the
syscall out of line and provide its own unwind information. */
# undef INTERNAL_SYSCALL_RAW
# define INTERNAL_SYSCALL_RAW(name, nr, args...) \
({ \
register int _a1 asm ("a1"); \
int _nametmp = name; \
LOAD_ARGS_##nr (args) \
register int _name asm ("ip") = _nametmp; \
asm volatile ("bl __libc_do_syscall" \
: "=r" (_a1) \
: "r" (_name) ASM_ARGS_##nr \
: "memory", "lr"); \
_a1; })
#else /* ARM */
# undef INTERNAL_SYSCALL_RAW
# define INTERNAL_SYSCALL_RAW(name, nr, args...) \
({ \
register int _a1 asm ("r0"), _nr asm ("r7"); \
LOAD_ARGS_##nr (args) \
_nr = name; \
asm volatile ("swi 0x0 @ syscall " #name \
: "=r" (_a1) \
: "r" (_nr) ASM_ARGS_##nr \
: "memory"); \
_a1; })
#endif
#undef INTERNAL_SYSCALL
#define INTERNAL_SYSCALL(name, nr, args...) \
INTERNAL_SYSCALL_RAW(SYS_ify(name), nr, args)
#define VDSO_NAME "LINUX_2.6"
#define VDSO_HASH 61765110
/* List of system calls which are supported as vsyscalls. */
#define HAVE_CLOCK_GETTIME_VSYSCALL "__vdso_clock_gettime"
#define HAVE_CLOCK_GETTIME64_VSYSCALL "__vdso_clock_gettime64"
#define HAVE_GETTIMEOFDAY_VSYSCALL "__vdso_gettimeofday"
#define LOAD_ARGS_0()
#define ASM_ARGS_0
#define LOAD_ARGS_1(a1) \
int _a1tmp = (int) (a1); \
LOAD_ARGS_0 () \
_a1 = _a1tmp;
#define ASM_ARGS_1 ASM_ARGS_0, "r" (_a1)
#define LOAD_ARGS_2(a1, a2) \
int _a2tmp = (int) (a2); \
LOAD_ARGS_1 (a1) \
register int _a2 asm ("a2") = _a2tmp;
#define ASM_ARGS_2 ASM_ARGS_1, "r" (_a2)
#define LOAD_ARGS_3(a1, a2, a3) \
int _a3tmp = (int) (a3); \
LOAD_ARGS_2 (a1, a2) \
register int _a3 asm ("a3") = _a3tmp;
#define ASM_ARGS_3 ASM_ARGS_2, "r" (_a3)
#define LOAD_ARGS_4(a1, a2, a3, a4) \
int _a4tmp = (int) (a4); \
LOAD_ARGS_3 (a1, a2, a3) \
register int _a4 asm ("a4") = _a4tmp;
#define ASM_ARGS_4 ASM_ARGS_3, "r" (_a4)
#define LOAD_ARGS_5(a1, a2, a3, a4, a5) \
int _v1tmp = (int) (a5); \
LOAD_ARGS_4 (a1, a2, a3, a4) \
register int _v1 asm ("v1") = _v1tmp;
#define ASM_ARGS_5 ASM_ARGS_4, "r" (_v1)
#define LOAD_ARGS_6(a1, a2, a3, a4, a5, a6) \
int _v2tmp = (int) (a6); \
LOAD_ARGS_5 (a1, a2, a3, a4, a5) \
register int _v2 asm ("v2") = _v2tmp;
#define ASM_ARGS_6 ASM_ARGS_5, "r" (_v2)
#ifndef __thumb__
# define LOAD_ARGS_7(a1, a2, a3, a4, a5, a6, a7) \
int _v3tmp = (int) (a7); \
LOAD_ARGS_6 (a1, a2, a3, a4, a5, a6) \
register int _v3 asm ("v3") = _v3tmp;
# define ASM_ARGS_7 ASM_ARGS_6, "r" (_v3)
#endif
/* For EABI, non-constant syscalls are actually pretty easy... */
#undef INTERNAL_SYSCALL_NCS
#define INTERNAL_SYSCALL_NCS(number, nr, args...) \
INTERNAL_SYSCALL_RAW (number, nr, args)
#define SINGLE_THREAD_BY_GLOBAL 1
#endif /* __ASSEMBLER__ */
#endif /* linux/arm/sysdep.h */