cosmopolitan/libc/runtime/cosmo.S

Ignoring revisions in .git-blame-ignore-revs. Click here to bypass and see the normal blame view.

149 lines
5.1 KiB
ArmAsm
Raw Permalink Normal View History

/*-*- mode:unix-assembly; indent-tabs-mode:t; tab-width:8; coding:utf-8 -*-│
2020-06-15 14:18:57 +00:00
vi: set noet ft=asm ts=8 sw=8 fenc=utf-8 :vi
Copyright 2020 Justine Alexandra Roberts Tunney
2020-12-28 01:18:44 +00:00
Permission to use, copy, modify, and/or distribute this software for
any purpose with or without fee is hereby granted, provided that the
above copyright notice and this permission notice appear in all copies.
2020-06-15 14:18:57 +00:00
2020-12-28 01:18:44 +00:00
THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL
WARRANTIES WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED
WARRANTIES OF MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE
AUTHOR BE LIABLE FOR ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL
DAMAGES OR ANY DAMAGES WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR
PROFITS, WHETHER IN AN ACTION OF CONTRACT, NEGLIGENCE OR OTHER
TORTIOUS ACTION, ARISING OUT OF OR IN CONNECTION WITH THE USE OR
PERFORMANCE OF THIS SOFTWARE.
2020-06-15 14:18:57 +00:00
*/
#include "libc/macros.h"
#include "libc/sysv/consts/prot.h"
#include "libc/sysv/consts/map.h"
#include "libc/intrin/strace.h"
#include "libc/sysv/consts/map.h"
2020-06-15 14:18:57 +00:00
#include "libc/dce.h"
#ifdef __x86_64__
2020-06-15 14:18:57 +00:00
.text.startup
// Cosmopolitan runtime.
//
// @param edi is argc
// @param rsi is argv
// @param rdx is environ
// @param rcx is auxv
// @noreturn
cosmo: push %rbp
mov %rsp,%rbp
mov %edi,%r12d
mov %rsi,%r13
mov %rdx,%r14
mov %rcx,%r15
#if SYSDEBUG
call __strace_init
mov %eax,%r12d
#endif /* SYSDEBUG */
#ifndef NOX87
// Windows always initializes FPU to double precision.
// WSL breaks Linux ABI by initializing FPU to double precision.
// This code makes long double long again.
//
// @see System V Application Binary Interface NexGen32e Architecture
// Processor Supplement, Version 1.0, December 5th, 2018
// Section 3.4.1: Initial Stack and Register State
fldcw 1f(%rip)
.rodata
2023-05-02 02:43:59 +00:00
.balign 2
// 8087 FPU Control Word
// IM: Invalid Operation
// DM: Denormal Operand
// ZM: Zero Divide
// OM: Overflow
// UM: Underflow
// PM: Precision
// PC: Precision Control
// {float,,double,long double}
// RC: Rounding Control
// {even, -, +, 0}
//
// drr
1: .short 0b00000000000000000001101111111
.previous
#endif
#ifdef __FAST_MATH__
2021-02-08 12:04:42 +00:00
push %rax
stmxcsr (%rsp)
//
// Enable hardware optimizations in violation of the IEEE standard.
//
// - 0x0040 enables "DAZ: Denormals Are Zeros" in MXCSR. This causes the
// processor to turn denormal inputs into zero, before computing them.
// See Intel Manual Vol. 1 §10.2.3.4
//
// - 0x8000 enables "FTZ: Flush To Zero" in MXCSR. This means a floating
// point operation that results in underflow will be set to zero, with
// the same sign, rather than producing a denormalized output. It will
// happen only if underflow trapping hasnt been enabled. See the Intel
// Manual Vol. 1 §10.2.3.3.
//
2021-02-08 12:04:42 +00:00
orl $0x8040,(%rsp)
ldmxcsr (%rsp)
pop %rax
#endif
// run assembly init
call _init
// call constructors
Release Cosmopolitan v3.3 This change upgrades to GCC 12.3 and GNU binutils 2.42. The GNU linker appears to have changed things so that only a single de-duplicated str table is present in the binary, and it gets placed wherever the linker wants, regardless of what the linker script says. To cope with that we need to stop using .ident to embed licenses. As such, this change does significant work to revamp how third party licenses are defined in the codebase, using `.section .notice,"aR",@progbits`. This new GCC 12.3 toolchain has support for GNU indirect functions. It lets us support __target_clones__ for the first time. This is used for optimizing the performance of libc string functions such as strlen and friends so far on x86, by ensuring AVX systems favor a second codepath that uses VEX encoding. It shaves some latency off certain operations. It's a useful feature to have for scientific computing for the reasons explained by the test/libcxx/openmp_test.cc example which compiles for fifteen different microarchitectures. Thanks to the upgrades, it's now also possible to use newer instruction sets, such as AVX512FP16, VNNI. Cosmo now uses the %gs register on x86 by default for TLS. Doing it is helpful for any program that links `cosmo_dlopen()`. Such programs had to recompile their binaries at startup to change the TLS instructions. That's not great, since it means every page in the executable needs to be faulted. The work of rewriting TLS-related x86 opcodes, is moved to fixupobj.com instead. This is great news for MacOS x86 users, since we previously needed to morph the binary every time for that platform but now that's no longer necessary. The only platforms where we need fixup of TLS x86 opcodes at runtime are now Windows, OpenBSD, and NetBSD. On Windows we morph TLS to point deeper into the TIB, based on a TlsAlloc assignment, and on OpenBSD/NetBSD we morph %gs back into %fs since the kernels do not allow us to specify a value for the %gs register. OpenBSD users are now required to use APE Loader to run Cosmo binaries and assimilation is no longer possible. OpenBSD kernel needs to change to allow programs to specify a value for the %gs register, or it needs to stop marking executable pages loaded by the kernel as mimmutable(). This release fixes __constructor__, .ctor, .init_array, and lastly the .preinit_array so they behave the exact same way as glibc. We no longer use hex constants to define math.h symbols like M_PI.
2024-02-20 19:12:09 +00:00
.weak __init_array_end
.weak __init_array_start
mov $__init_array_start,%eax
1: cmp $__init_array_end,%eax
2021-02-08 12:04:42 +00:00
je 2f
push %rax
Release Cosmopolitan v3.3 This change upgrades to GCC 12.3 and GNU binutils 2.42. The GNU linker appears to have changed things so that only a single de-duplicated str table is present in the binary, and it gets placed wherever the linker wants, regardless of what the linker script says. To cope with that we need to stop using .ident to embed licenses. As such, this change does significant work to revamp how third party licenses are defined in the codebase, using `.section .notice,"aR",@progbits`. This new GCC 12.3 toolchain has support for GNU indirect functions. It lets us support __target_clones__ for the first time. This is used for optimizing the performance of libc string functions such as strlen and friends so far on x86, by ensuring AVX systems favor a second codepath that uses VEX encoding. It shaves some latency off certain operations. It's a useful feature to have for scientific computing for the reasons explained by the test/libcxx/openmp_test.cc example which compiles for fifteen different microarchitectures. Thanks to the upgrades, it's now also possible to use newer instruction sets, such as AVX512FP16, VNNI. Cosmo now uses the %gs register on x86 by default for TLS. Doing it is helpful for any program that links `cosmo_dlopen()`. Such programs had to recompile their binaries at startup to change the TLS instructions. That's not great, since it means every page in the executable needs to be faulted. The work of rewriting TLS-related x86 opcodes, is moved to fixupobj.com instead. This is great news for MacOS x86 users, since we previously needed to morph the binary every time for that platform but now that's no longer necessary. The only platforms where we need fixup of TLS x86 opcodes at runtime are now Windows, OpenBSD, and NetBSD. On Windows we morph TLS to point deeper into the TIB, based on a TlsAlloc assignment, and on OpenBSD/NetBSD we morph %gs back into %fs since the kernels do not allow us to specify a value for the %gs register. OpenBSD users are now required to use APE Loader to run Cosmo binaries and assimilation is no longer possible. OpenBSD kernel needs to change to allow programs to specify a value for the %gs register, or it needs to stop marking executable pages loaded by the kernel as mimmutable(). This release fixes __constructor__, .ctor, .init_array, and lastly the .preinit_array so they behave the exact same way as glibc. We no longer use hex constants to define math.h symbols like M_PI.
2024-02-20 19:12:09 +00:00
push %rax
2021-02-08 12:04:42 +00:00
call .Largs
call *(%rax)
pop %rax
Release Cosmopolitan v3.3 This change upgrades to GCC 12.3 and GNU binutils 2.42. The GNU linker appears to have changed things so that only a single de-duplicated str table is present in the binary, and it gets placed wherever the linker wants, regardless of what the linker script says. To cope with that we need to stop using .ident to embed licenses. As such, this change does significant work to revamp how third party licenses are defined in the codebase, using `.section .notice,"aR",@progbits`. This new GCC 12.3 toolchain has support for GNU indirect functions. It lets us support __target_clones__ for the first time. This is used for optimizing the performance of libc string functions such as strlen and friends so far on x86, by ensuring AVX systems favor a second codepath that uses VEX encoding. It shaves some latency off certain operations. It's a useful feature to have for scientific computing for the reasons explained by the test/libcxx/openmp_test.cc example which compiles for fifteen different microarchitectures. Thanks to the upgrades, it's now also possible to use newer instruction sets, such as AVX512FP16, VNNI. Cosmo now uses the %gs register on x86 by default for TLS. Doing it is helpful for any program that links `cosmo_dlopen()`. Such programs had to recompile their binaries at startup to change the TLS instructions. That's not great, since it means every page in the executable needs to be faulted. The work of rewriting TLS-related x86 opcodes, is moved to fixupobj.com instead. This is great news for MacOS x86 users, since we previously needed to morph the binary every time for that platform but now that's no longer necessary. The only platforms where we need fixup of TLS x86 opcodes at runtime are now Windows, OpenBSD, and NetBSD. On Windows we morph TLS to point deeper into the TIB, based on a TlsAlloc assignment, and on OpenBSD/NetBSD we morph %gs back into %fs since the kernels do not allow us to specify a value for the %gs register. OpenBSD users are now required to use APE Loader to run Cosmo binaries and assimilation is no longer possible. OpenBSD kernel needs to change to allow programs to specify a value for the %gs register, or it needs to stop marking executable pages loaded by the kernel as mimmutable(). This release fixes __constructor__, .ctor, .init_array, and lastly the .preinit_array so they behave the exact same way as glibc. We no longer use hex constants to define math.h symbols like M_PI.
2024-02-20 19:12:09 +00:00
pop %rax
add $8,%eax
2021-02-08 12:04:42 +00:00
jmp 1b
// call main()
Make numerous improvements - Python static hello world now 1.8mb - Python static fully loaded now 10mb - Python HTTPS client now uses MbedTLS - Python REPL now completes import stmts - Increase stack size for Python for now - Begin synthesizing posixpath and ntpath - Restore Python \N{UNICODE NAME} support - Restore Python NFKD symbol normalization - Add optimized code path for Intel SHA-NI - Get more Python unit tests passing faster - Get Python help() pagination working on NT - Python hashlib now supports MbedTLS PBKDF2 - Make memcpy/memmove/memcmp/bcmp/etc. faster - Add Mersenne Twister and Vigna to LIBC_RAND - Provide privileged __printf() for error code - Fix zipos opendir() so that it reports ENOTDIR - Add basic chmod() implementation for Windows NT - Add Cosmo's best functions to Python cosmo module - Pin function trace indent depth to that of caller - Show memory diagram on invalid access in MODE=dbg - Differentiate stack overflow on crash in MODE=dbg - Add stb_truetype and tools for analyzing font files - Upgrade to UNICODE 13 and reduce its binary footprint - COMPILE.COM now logs resource usage of build commands - Start implementing basic poll() support on bare metal - Set getauxval(AT_EXECFN) to GetModuleFileName() on NT - Add descriptions to strerror() in non-TINY build modes - Add COUNTBRANCH() macro to help with micro-optimizations - Make error / backtrace / asan / memory code more unbreakable - Add fast perfect C implementation of μ-Law and a-Law audio codecs - Make strtol() functions consistent with other libc implementations - Improve Linenoise implementation (see also github.com/jart/bestline) - COMPILE.COM now suppresses stdout/stderr of successful build commands
2021-09-28 05:58:51 +00:00
2: call .Largs
2020-06-15 14:18:57 +00:00
.weak main
call main
xchg %eax,%edi
2020-06-15 14:18:57 +00:00
call exit
2021-02-08 12:04:42 +00:00
.Largs: mov %r12d,%edi
mov %r13,%rsi
mov %r14,%rdx
mov %r15,%rcx
ret
.endfn cosmo,weak
2020-06-15 14:18:57 +00:00
// Enables Thread Local Storage.
.init.start 304,_init_tls
push %rdi
push %rsi
call __enable_tls
pop %rsi
pop %rdi
.init.end 304,_init_tls
#ifdef FTRACE
2022-06-13 05:20:59 +00:00
.init.start 306,_init_ftrace
push %rdi
push %rsi
2022-06-13 05:20:59 +00:00
call ftrace_init
mov %eax,%r12d
pop %rsi
pop %rdi
2022-06-13 05:20:59 +00:00
.init.end 306,_init_ftrace
#endif
#endif /* __x86_64__ */