1
0
mirror of https://github.com/golang/go synced 2024-10-02 08:28:36 -06:00
go/src/pkg/runtime/asm_arm.s

513 lines
13 KiB
ArmAsm
Raw Normal View History

// Copyright 2009 The Go Authors. All rights reserved.
// Use of this source code is governed by a BSD-style
// license that can be found in the LICENSE file.
#include "zasm_GOOS_GOARCH.h"
// using frame size $-4 means do not save LR on stack.
TEXT _rt0_go(SB),7,$-4
MOVW $0xcafebabe, R12
// copy arguments forward on an even stack
// use R13 instead of SP to avoid linker rewriting the offsets
MOVW 0(R13), R0 // argc
MOVW $4(R13), R1 // argv
ld: detect stack overflow due to NOSPLIT Fix problems found. On amd64, various library routines had bigger stack frames than expected, because large function calls had been added. runtime.assertI2T: nosplit stack overflow 120 assumed on entry to runtime.assertI2T 8 after runtime.assertI2T uses 112 0 on entry to runtime.newTypeAssertionError -8 on entry to runtime.morestack01 runtime.assertE2E: nosplit stack overflow 120 assumed on entry to runtime.assertE2E 16 after runtime.assertE2E uses 104 8 on entry to runtime.panic 0 on entry to runtime.morestack16 -8 after runtime.morestack16 uses 8 runtime.assertE2T: nosplit stack overflow 120 assumed on entry to runtime.assertE2T 16 after runtime.assertE2T uses 104 8 on entry to runtime.panic 0 on entry to runtime.morestack16 -8 after runtime.morestack16 uses 8 runtime.newselect: nosplit stack overflow 120 assumed on entry to runtime.newselect 56 after runtime.newselect uses 64 48 on entry to runtime.printf 8 after runtime.printf uses 40 0 on entry to vprintf -8 on entry to runtime.morestack16 runtime.selectdefault: nosplit stack overflow 120 assumed on entry to runtime.selectdefault 56 after runtime.selectdefault uses 64 48 on entry to runtime.printf 8 after runtime.printf uses 40 0 on entry to vprintf -8 on entry to runtime.morestack16 runtime.selectgo: nosplit stack overflow 120 assumed on entry to runtime.selectgo 0 after runtime.selectgo uses 120 -8 on entry to runtime.gosched On arm, 5c was tagging functions NOSPLIT that should not have been, like the recursive function printpanics: printpanics: nosplit stack overflow 124 assumed on entry to printpanics 112 after printpanics uses 12 108 on entry to printpanics 96 after printpanics uses 12 92 on entry to printpanics 80 after printpanics uses 12 76 on entry to printpanics 64 after printpanics uses 12 60 on entry to printpanics 48 after printpanics uses 12 44 on entry to printpanics 32 after printpanics uses 12 28 on entry to printpanics 16 after printpanics uses 12 12 on entry to printpanics 0 after printpanics uses 12 -4 on entry to printpanics R=r, r2 CC=golang-dev https://golang.org/cl/4188061
2011-02-22 15:40:40 -07:00
SUB $64, R13 // plenty of scratch
AND $~7, R13
ld: detect stack overflow due to NOSPLIT Fix problems found. On amd64, various library routines had bigger stack frames than expected, because large function calls had been added. runtime.assertI2T: nosplit stack overflow 120 assumed on entry to runtime.assertI2T 8 after runtime.assertI2T uses 112 0 on entry to runtime.newTypeAssertionError -8 on entry to runtime.morestack01 runtime.assertE2E: nosplit stack overflow 120 assumed on entry to runtime.assertE2E 16 after runtime.assertE2E uses 104 8 on entry to runtime.panic 0 on entry to runtime.morestack16 -8 after runtime.morestack16 uses 8 runtime.assertE2T: nosplit stack overflow 120 assumed on entry to runtime.assertE2T 16 after runtime.assertE2T uses 104 8 on entry to runtime.panic 0 on entry to runtime.morestack16 -8 after runtime.morestack16 uses 8 runtime.newselect: nosplit stack overflow 120 assumed on entry to runtime.newselect 56 after runtime.newselect uses 64 48 on entry to runtime.printf 8 after runtime.printf uses 40 0 on entry to vprintf -8 on entry to runtime.morestack16 runtime.selectdefault: nosplit stack overflow 120 assumed on entry to runtime.selectdefault 56 after runtime.selectdefault uses 64 48 on entry to runtime.printf 8 after runtime.printf uses 40 0 on entry to vprintf -8 on entry to runtime.morestack16 runtime.selectgo: nosplit stack overflow 120 assumed on entry to runtime.selectgo 0 after runtime.selectgo uses 120 -8 on entry to runtime.gosched On arm, 5c was tagging functions NOSPLIT that should not have been, like the recursive function printpanics: printpanics: nosplit stack overflow 124 assumed on entry to printpanics 112 after printpanics uses 12 108 on entry to printpanics 96 after printpanics uses 12 92 on entry to printpanics 80 after printpanics uses 12 76 on entry to printpanics 64 after printpanics uses 12 60 on entry to printpanics 48 after printpanics uses 12 44 on entry to printpanics 32 after printpanics uses 12 28 on entry to printpanics 16 after printpanics uses 12 12 on entry to printpanics 0 after printpanics uses 12 -4 on entry to printpanics R=r, r2 CC=golang-dev https://golang.org/cl/4188061
2011-02-22 15:40:40 -07:00
MOVW R0, 60(R13) // save argc, argv away
MOVW R1, 64(R13)
// set up m and g registers
// g is R10, m is R9
MOVW $runtime·g0(SB), g
MOVW $runtime·m0(SB), m
// save m->g0 = g0
MOVW g, m_g0(m)
// create istack out of the OS stack
MOVW $(-8192+104)(R13), R0
MOVW R0, g_stackguard(g) // (w 104b guard)
MOVW R0, g_stackguard0(g)
MOVW R13, g_stackbase(g)
BL runtime·emptyfunc(SB) // fault if stack check is wrong
// if there is an _cgo_init, call it.
MOVW _cgo_init(SB), R2
CMP $0, R2
MOVW.NE g, R0 // first argument of _cgo_init is g
BL.NE (R2) // will clobber R0-R3
// update stackguard after _cgo_init
MOVW g_stackguard0(g), R0
MOVW R0, g_stackguard(g)
BL runtime·checkgoarm(SB)
BL runtime·check(SB)
// saved argc, argv
ld: detect stack overflow due to NOSPLIT Fix problems found. On amd64, various library routines had bigger stack frames than expected, because large function calls had been added. runtime.assertI2T: nosplit stack overflow 120 assumed on entry to runtime.assertI2T 8 after runtime.assertI2T uses 112 0 on entry to runtime.newTypeAssertionError -8 on entry to runtime.morestack01 runtime.assertE2E: nosplit stack overflow 120 assumed on entry to runtime.assertE2E 16 after runtime.assertE2E uses 104 8 on entry to runtime.panic 0 on entry to runtime.morestack16 -8 after runtime.morestack16 uses 8 runtime.assertE2T: nosplit stack overflow 120 assumed on entry to runtime.assertE2T 16 after runtime.assertE2T uses 104 8 on entry to runtime.panic 0 on entry to runtime.morestack16 -8 after runtime.morestack16 uses 8 runtime.newselect: nosplit stack overflow 120 assumed on entry to runtime.newselect 56 after runtime.newselect uses 64 48 on entry to runtime.printf 8 after runtime.printf uses 40 0 on entry to vprintf -8 on entry to runtime.morestack16 runtime.selectdefault: nosplit stack overflow 120 assumed on entry to runtime.selectdefault 56 after runtime.selectdefault uses 64 48 on entry to runtime.printf 8 after runtime.printf uses 40 0 on entry to vprintf -8 on entry to runtime.morestack16 runtime.selectgo: nosplit stack overflow 120 assumed on entry to runtime.selectgo 0 after runtime.selectgo uses 120 -8 on entry to runtime.gosched On arm, 5c was tagging functions NOSPLIT that should not have been, like the recursive function printpanics: printpanics: nosplit stack overflow 124 assumed on entry to printpanics 112 after printpanics uses 12 108 on entry to printpanics 96 after printpanics uses 12 92 on entry to printpanics 80 after printpanics uses 12 76 on entry to printpanics 64 after printpanics uses 12 60 on entry to printpanics 48 after printpanics uses 12 44 on entry to printpanics 32 after printpanics uses 12 28 on entry to printpanics 16 after printpanics uses 12 12 on entry to printpanics 0 after printpanics uses 12 -4 on entry to printpanics R=r, r2 CC=golang-dev https://golang.org/cl/4188061
2011-02-22 15:40:40 -07:00
MOVW 60(R13), R0
MOVW R0, 4(R13)
ld: detect stack overflow due to NOSPLIT Fix problems found. On amd64, various library routines had bigger stack frames than expected, because large function calls had been added. runtime.assertI2T: nosplit stack overflow 120 assumed on entry to runtime.assertI2T 8 after runtime.assertI2T uses 112 0 on entry to runtime.newTypeAssertionError -8 on entry to runtime.morestack01 runtime.assertE2E: nosplit stack overflow 120 assumed on entry to runtime.assertE2E 16 after runtime.assertE2E uses 104 8 on entry to runtime.panic 0 on entry to runtime.morestack16 -8 after runtime.morestack16 uses 8 runtime.assertE2T: nosplit stack overflow 120 assumed on entry to runtime.assertE2T 16 after runtime.assertE2T uses 104 8 on entry to runtime.panic 0 on entry to runtime.morestack16 -8 after runtime.morestack16 uses 8 runtime.newselect: nosplit stack overflow 120 assumed on entry to runtime.newselect 56 after runtime.newselect uses 64 48 on entry to runtime.printf 8 after runtime.printf uses 40 0 on entry to vprintf -8 on entry to runtime.morestack16 runtime.selectdefault: nosplit stack overflow 120 assumed on entry to runtime.selectdefault 56 after runtime.selectdefault uses 64 48 on entry to runtime.printf 8 after runtime.printf uses 40 0 on entry to vprintf -8 on entry to runtime.morestack16 runtime.selectgo: nosplit stack overflow 120 assumed on entry to runtime.selectgo 0 after runtime.selectgo uses 120 -8 on entry to runtime.gosched On arm, 5c was tagging functions NOSPLIT that should not have been, like the recursive function printpanics: printpanics: nosplit stack overflow 124 assumed on entry to printpanics 112 after printpanics uses 12 108 on entry to printpanics 96 after printpanics uses 12 92 on entry to printpanics 80 after printpanics uses 12 76 on entry to printpanics 64 after printpanics uses 12 60 on entry to printpanics 48 after printpanics uses 12 44 on entry to printpanics 32 after printpanics uses 12 28 on entry to printpanics 16 after printpanics uses 12 12 on entry to printpanics 0 after printpanics uses 12 -4 on entry to printpanics R=r, r2 CC=golang-dev https://golang.org/cl/4188061
2011-02-22 15:40:40 -07:00
MOVW 64(R13), R1
MOVW R1, 8(R13)
BL runtime·args(SB)
BL runtime·osinit(SB)
BL runtime·hashinit(SB)
BL runtime·schedinit(SB)
// create a new goroutine to start program
MOVW $runtime·main·f(SB), R0
MOVW.W R0, -4(R13)
MOVW $8, R0
MOVW.W R0, -4(R13)
MOVW $0, R0
MOVW.W R0, -4(R13) // push $0 as guard
BL runtime·newproc(SB)
MOVW $12(R13), R13 // pop args and LR
// start this M
BL runtime·mstart(SB)
MOVW $1234, R0
MOVW $1000, R1
MOVW R0, (R1) // fail hard
DATA runtime·main·f+0(SB)/4,$runtime·main(SB)
GLOBL runtime·main·f(SB),8,$4
TEXT runtime·breakpoint(SB),7,$0
// gdb won't skip this breakpoint instruction automatically,
// so you must manually "set $pc+=4" to skip it and continue.
WORD $0xe1200071 // BKPT 0x0001
RET
GLOBL runtime·goarm(SB), $4
TEXT runtime·asminit(SB),7,$0
// disable runfast (flush-to-zero) mode of vfp if runtime.goarm > 5
MOVW runtime·goarm(SB), R11
CMP $5, R11
BLE 4(PC)
WORD $0xeef1ba10 // vmrs r11, fpscr
BIC $(1<<24), R11
WORD $0xeee1ba10 // vmsr fpscr, r11
RET
/*
* go-routine
*/
runtime: scheduler, cgo reorganization * Change use of m->g0 stack (aka scheduler stack). * Provide runtime.mcall(f) to invoke f() on m->g0 stack. * Replace scheduler loop entry with runtime.mcall(schedule). Runtime.mcall eliminates the need for fake scheduler states that exist just to run a bit of code on the m->g0 stack (Grecovery, Gstackalloc). The elimination of the scheduler as a loop that stops and starts using gosave and gogo fixes a bad interaction with the way cgo uses the m->g0 stack. Cgo runs external (gcc-compiled) C functions on that stack, and then when calling back into Go, it sets m->g0->sched.sp below the added call frames, so that other uses of m->g0's stack will not interfere with those frames. Unfortunately, gogo (longjmp) back to the scheduler loop at this point would end up running scheduler with the lower sp, which no longer points at a valid stack frame for a call to scheduler. If scheduler then wrote any function call arguments or local variables to where it expected the stack frame to be, it would overwrite other data on the stack. I realized this possibility while debugging a problem with calling complex Go code in a Go -> C -> Go cgo callback. This wasn't the bug I was looking for, it turns out, but I believe it is a real bug nonetheless. Switching to runtime.mcall, which only adds new frames to the stack and never jumps into functions running in existing ones, fixes this bug. * Move cgo-related code out of proc.c into cgocall.c. * Add very large comment describing cgo call sequences. * Simpilify, regularize cgo function implementations and names. * Add test suite as misc/cgo/test. Now the Go -> C path calls cgocall, which calls asmcgocall, and the C -> Go path calls cgocallback, which calls cgocallbackg. The shuffling, which affects mainly the callback case, moves most of the callback implementation to cgocallback running on the m->curg stack (not the m->g0 scheduler stack) and only while accounted for with $GOMAXPROCS (between calls to exitsyscall and entersyscall). The previous callback code did not block in startcgocallback's approximation to exitsyscall, so if, say, the garbage collector were running, it would still barge in and start doing things like call malloc. Similarly endcgocallback's approximation of entersyscall did not call matchmg to kick off new OS threads when necessary, which caused the bug in issue 1560. Fixes #1560. R=iant CC=golang-dev https://golang.org/cl/4253054
2011-03-07 08:37:42 -07:00
// void gosave(Gobuf*)
// save state in Gobuf; setjmp
TEXT runtime·gosave(SB), 7, $-4
MOVW 0(FP), R0 // gobuf
MOVW SP, gobuf_sp(R0)
MOVW LR, gobuf_pc(R0)
MOVW g, gobuf_g(R0)
MOVW $0, R11
MOVW R11, gobuf_lr(R0)
MOVW R11, gobuf_ret(R0)
MOVW R11, gobuf_ctxt(R0)
RET
// void gogo(Gobuf*)
// restore state from Gobuf; longjmp
TEXT runtime·gogo(SB), 7, $-4
MOVW 0(FP), R1 // gobuf
MOVW gobuf_g(R1), g
MOVW 0(g), R2 // make sure g != nil
MOVW _cgo_save_gm(SB), R2
CMP $0, R2 // if in Cgo, we have to save g and m
BL.NE (R2) // this call will clobber R0
MOVW gobuf_sp(R1), SP // restore SP
MOVW gobuf_lr(R1), LR
MOVW gobuf_ret(R1), R0
MOVW gobuf_ctxt(R1), R7
MOVW $0, R11
MOVW R11, gobuf_sp(R1) // clear to help garbage collector
MOVW R11, gobuf_ret(R1)
MOVW R11, gobuf_lr(R1)
MOVW R11, gobuf_ctxt(R1)
runtime: record proper goroutine state during stack split Until now, the goroutine state has been scattered during the execution of newstack and oldstack. It's all there, and those routines know how to get back to a working goroutine, but other pieces of the system, like stack traces, do not. If something does interrupt the newstack or oldstack execution, the rest of the system can't understand the goroutine. For example, if newstack decides there is an overflow and calls throw, the stack tracer wouldn't dump the goroutine correctly. For newstack to save a useful state snapshot, it needs to be able to rewind the PC in the function that triggered the split back to the beginning of the function. (The PC is a few instructions in, just after the call to morestack.) To make that possible, we change the prologues to insert a jmp back to the beginning of the function after the call to morestack. That is, the prologue used to be roughly: TEXT myfunc check for split jmpcond nosplit call morestack nosplit: sub $xxx, sp Now an extra instruction is inserted after the call: TEXT myfunc start: check for split jmpcond nosplit call morestack jmp start nosplit: sub $xxx, sp The jmp is not executed directly. It is decoded and simulated by runtime.rewindmorestack to discover the beginning of the function, and then the call to morestack returns directly to the start label instead of to the jump instruction. So logically the jmp is still executed, just not by the cpu. The prologue thus repeats in the case of a function that needs a stack split, but against the cost of the split itself, the extra few instructions are noise. The repeated prologue has the nice effect of making a stack split double-check that the new stack is big enough: if morestack happens to return on a too-small stack, we'll now notice before corruption happens. The ability for newstack to rewind to the beginning of the function should help preemption too. If newstack decides that it was called for preemption instead of a stack split, it now has the goroutine state correctly paused if rescheduling is needed, and when the goroutine can run again, it can return to the start label on its original stack and re-execute the split check. Here is an example of a split stack overflow showing the full trace, without any special cases in the stack printer. (This one was triggered by making the split check incorrect.) runtime: newstack framesize=0x0 argsize=0x18 sp=0x6aebd0 stack=[0x6b0000, 0x6b0fa0] morebuf={pc:0x69f5b sp:0x6aebd8 lr:0x0} sched={pc:0x68880 sp:0x6aebd0 lr:0x0 ctxt:0x34e700} runtime: split stack overflow: 0x6aebd0 < 0x6b0000 fatal error: runtime: split stack overflow goroutine 1 [stack split]: runtime.mallocgc(0x290, 0x100000000, 0x1) /Users/rsc/g/go/src/pkg/runtime/zmalloc_darwin_amd64.c:21 fp=0x6aebd8 runtime.new() /Users/rsc/g/go/src/pkg/runtime/zmalloc_darwin_amd64.c:682 +0x5b fp=0x6aec08 go/build.(*Context).Import(0x5ae340, 0xc210030c71, 0xa, 0xc2100b4380, 0x1b, ...) /Users/rsc/g/go/src/pkg/go/build/build.go:424 +0x3a fp=0x6b00a0 main.loadImport(0xc210030c71, 0xa, 0xc2100b4380, 0x1b, 0xc2100b42c0, ...) /Users/rsc/g/go/src/cmd/go/pkg.go:249 +0x371 fp=0x6b01a8 main.(*Package).load(0xc21017c800, 0xc2100b42c0, 0xc2101828c0, 0x0, 0x0, ...) /Users/rsc/g/go/src/cmd/go/pkg.go:431 +0x2801 fp=0x6b0c98 main.loadPackage(0x369040, 0x7, 0xc2100b42c0, 0x0) /Users/rsc/g/go/src/cmd/go/pkg.go:709 +0x857 fp=0x6b0f80 ----- stack segment boundary ----- main.(*builder).action(0xc2100902a0, 0x0, 0x0, 0xc2100e6c00, 0xc2100e5750, ...) /Users/rsc/g/go/src/cmd/go/build.go:539 +0x437 fp=0x6b14a0 main.(*builder).action(0xc2100902a0, 0x0, 0x0, 0xc21015b400, 0x2, ...) /Users/rsc/g/go/src/cmd/go/build.go:528 +0x1d2 fp=0x6b1658 main.(*builder).test(0xc2100902a0, 0xc210092000, 0x0, 0x0, 0xc21008ff60, ...) /Users/rsc/g/go/src/cmd/go/test.go:622 +0x1b53 fp=0x6b1f68 ----- stack segment boundary ----- main.runTest(0x5a6b20, 0xc21000a020, 0x2, 0x2) /Users/rsc/g/go/src/cmd/go/test.go:366 +0xd09 fp=0x6a5cf0 main.main() /Users/rsc/g/go/src/cmd/go/main.go:161 +0x4f9 fp=0x6a5f78 runtime.main() /Users/rsc/g/go/src/pkg/runtime/proc.c:183 +0x92 fp=0x6a5fa0 runtime.goexit() /Users/rsc/g/go/src/pkg/runtime/proc.c:1266 fp=0x6a5fa8 And here is a seg fault during oldstack: SIGSEGV: segmentation violation PC=0x1b2a6 runtime.oldstack() /Users/rsc/g/go/src/pkg/runtime/stack.c:159 +0x76 runtime.lessstack() /Users/rsc/g/go/src/pkg/runtime/asm_amd64.s:270 +0x22 goroutine 1 [stack unsplit]: fmt.(*pp).printArg(0x2102e64e0, 0xe5c80, 0x2102c9220, 0x73, 0x0, ...) /Users/rsc/g/go/src/pkg/fmt/print.go:818 +0x3d3 fp=0x221031e6f8 fmt.(*pp).doPrintf(0x2102e64e0, 0x12fb20, 0x2, 0x221031eb98, 0x1, ...) /Users/rsc/g/go/src/pkg/fmt/print.go:1183 +0x15cb fp=0x221031eaf0 fmt.Sprintf(0x12fb20, 0x2, 0x221031eb98, 0x1, 0x1, ...) /Users/rsc/g/go/src/pkg/fmt/print.go:234 +0x67 fp=0x221031eb40 flag.(*stringValue).String(0x2102c9210, 0x1, 0x0) /Users/rsc/g/go/src/pkg/flag/flag.go:180 +0xb3 fp=0x221031ebb0 flag.(*FlagSet).Var(0x2102f6000, 0x293d38, 0x2102c9210, 0x143490, 0xa, ...) /Users/rsc/g/go/src/pkg/flag/flag.go:633 +0x40 fp=0x221031eca0 flag.(*FlagSet).StringVar(0x2102f6000, 0x2102c9210, 0x143490, 0xa, 0x12fa60, ...) /Users/rsc/g/go/src/pkg/flag/flag.go:550 +0x91 fp=0x221031ece8 flag.(*FlagSet).String(0x2102f6000, 0x143490, 0xa, 0x12fa60, 0x0, ...) /Users/rsc/g/go/src/pkg/flag/flag.go:563 +0x87 fp=0x221031ed38 flag.String(0x143490, 0xa, 0x12fa60, 0x0, 0x161950, ...) /Users/rsc/g/go/src/pkg/flag/flag.go:570 +0x6b fp=0x221031ed80 testing.init() /Users/rsc/g/go/src/pkg/testing/testing.go:-531 +0xbb fp=0x221031edc0 strings_test.init() /Users/rsc/g/go/src/pkg/strings/strings_test.go:1115 +0x62 fp=0x221031ef70 main.init() strings/_test/_testmain.go:90 +0x3d fp=0x221031ef78 runtime.main() /Users/rsc/g/go/src/pkg/runtime/proc.c:180 +0x8a fp=0x221031efa0 runtime.goexit() /Users/rsc/g/go/src/pkg/runtime/proc.c:1269 fp=0x221031efa8 goroutine 2 [runnable]: runtime.MHeap_Scavenger() /Users/rsc/g/go/src/pkg/runtime/mheap.c:438 runtime.goexit() /Users/rsc/g/go/src/pkg/runtime/proc.c:1269 created by runtime.main /Users/rsc/g/go/src/pkg/runtime/proc.c:166 rax 0x23ccc0 rbx 0x23ccc0 rcx 0x0 rdx 0x38 rdi 0x2102c0170 rsi 0x221032cfe0 rbp 0x221032cfa0 rsp 0x7fff5fbff5b0 r8 0x2102c0120 r9 0x221032cfa0 r10 0x221032c000 r11 0x104ce8 r12 0xe5c80 r13 0x1be82baac718 r14 0x13091135f7d69200 r15 0x0 rip 0x1b2a6 rflags 0x10246 cs 0x2b fs 0x0 gs 0x0 Fixes #5723. R=r, dvyukov, go.peter.90, dave, iant CC=golang-dev https://golang.org/cl/10360048
2013-06-27 09:32:01 -06:00
CMP R11, R11 // set condition codes for == test, needed by stack split
MOVW gobuf_pc(R1), PC
runtime: scheduler, cgo reorganization * Change use of m->g0 stack (aka scheduler stack). * Provide runtime.mcall(f) to invoke f() on m->g0 stack. * Replace scheduler loop entry with runtime.mcall(schedule). Runtime.mcall eliminates the need for fake scheduler states that exist just to run a bit of code on the m->g0 stack (Grecovery, Gstackalloc). The elimination of the scheduler as a loop that stops and starts using gosave and gogo fixes a bad interaction with the way cgo uses the m->g0 stack. Cgo runs external (gcc-compiled) C functions on that stack, and then when calling back into Go, it sets m->g0->sched.sp below the added call frames, so that other uses of m->g0's stack will not interfere with those frames. Unfortunately, gogo (longjmp) back to the scheduler loop at this point would end up running scheduler with the lower sp, which no longer points at a valid stack frame for a call to scheduler. If scheduler then wrote any function call arguments or local variables to where it expected the stack frame to be, it would overwrite other data on the stack. I realized this possibility while debugging a problem with calling complex Go code in a Go -> C -> Go cgo callback. This wasn't the bug I was looking for, it turns out, but I believe it is a real bug nonetheless. Switching to runtime.mcall, which only adds new frames to the stack and never jumps into functions running in existing ones, fixes this bug. * Move cgo-related code out of proc.c into cgocall.c. * Add very large comment describing cgo call sequences. * Simpilify, regularize cgo function implementations and names. * Add test suite as misc/cgo/test. Now the Go -> C path calls cgocall, which calls asmcgocall, and the C -> Go path calls cgocallback, which calls cgocallbackg. The shuffling, which affects mainly the callback case, moves most of the callback implementation to cgocallback running on the m->curg stack (not the m->g0 scheduler stack) and only while accounted for with $GOMAXPROCS (between calls to exitsyscall and entersyscall). The previous callback code did not block in startcgocallback's approximation to exitsyscall, so if, say, the garbage collector were running, it would still barge in and start doing things like call malloc. Similarly endcgocallback's approximation of entersyscall did not call matchmg to kick off new OS threads when necessary, which caused the bug in issue 1560. Fixes #1560. R=iant CC=golang-dev https://golang.org/cl/4253054
2011-03-07 08:37:42 -07:00
// void mcall(void (*fn)(G*))
// Switch to m->g0's stack, call fn(g).
runtime: stack split + garbage collection bug The g->sched.sp saved stack pointer and the g->stackbase and g->stackguard stack bounds can change even while "the world is stopped", because a goroutine has to call functions (and therefore might split its stack) when exiting a system call to check whether the world is stopped (and if so, wait until the world continues). That means the garbage collector cannot access those values safely (without a race) for goroutines executing system calls. Instead, save a consistent triple in g->gcsp, g->gcstack, g->gcguard during entersyscall and have the garbage collector refer to those. The old code was occasionally seeing (because of the race) an sp and stk that did not correspond to each other, so that stk - sp was not the number of stack bytes following sp. In that case, if sp < stk then the call scanblock(sp, stk - sp) scanned too many bytes (anything between the two pointers, which pointed into different allocation blocks). If sp > stk then stk - sp wrapped around. On 32-bit, stk - sp is a uintptr (uint32) converted to int64 in the call to scanblock, so a large (~4G) but positive number. Scanblock would try to scan that many bytes and eventually fault accessing unmapped memory. On 64-bit, stk - sp is a uintptr (uint64) promoted to int64 in the call to scanblock, so a negative number. Scanblock would not scan anything, possibly causing in-use blocks to be freed. In short, 32-bit platforms would have seen either ineffective garbage collection or crashes during garbage collection, while 64-bit platforms would have seen either ineffective or incorrect garbage collection. You can see the invalid arguments to scanblock in the stack traces in issue 1620. Fixes #1620. Fixes #1746. R=iant, r CC=golang-dev https://golang.org/cl/4437075
2011-04-27 21:21:12 -06:00
// Fn must never return. It should gogo(&g->sched)
runtime: scheduler, cgo reorganization * Change use of m->g0 stack (aka scheduler stack). * Provide runtime.mcall(f) to invoke f() on m->g0 stack. * Replace scheduler loop entry with runtime.mcall(schedule). Runtime.mcall eliminates the need for fake scheduler states that exist just to run a bit of code on the m->g0 stack (Grecovery, Gstackalloc). The elimination of the scheduler as a loop that stops and starts using gosave and gogo fixes a bad interaction with the way cgo uses the m->g0 stack. Cgo runs external (gcc-compiled) C functions on that stack, and then when calling back into Go, it sets m->g0->sched.sp below the added call frames, so that other uses of m->g0's stack will not interfere with those frames. Unfortunately, gogo (longjmp) back to the scheduler loop at this point would end up running scheduler with the lower sp, which no longer points at a valid stack frame for a call to scheduler. If scheduler then wrote any function call arguments or local variables to where it expected the stack frame to be, it would overwrite other data on the stack. I realized this possibility while debugging a problem with calling complex Go code in a Go -> C -> Go cgo callback. This wasn't the bug I was looking for, it turns out, but I believe it is a real bug nonetheless. Switching to runtime.mcall, which only adds new frames to the stack and never jumps into functions running in existing ones, fixes this bug. * Move cgo-related code out of proc.c into cgocall.c. * Add very large comment describing cgo call sequences. * Simpilify, regularize cgo function implementations and names. * Add test suite as misc/cgo/test. Now the Go -> C path calls cgocall, which calls asmcgocall, and the C -> Go path calls cgocallback, which calls cgocallbackg. The shuffling, which affects mainly the callback case, moves most of the callback implementation to cgocallback running on the m->curg stack (not the m->g0 scheduler stack) and only while accounted for with $GOMAXPROCS (between calls to exitsyscall and entersyscall). The previous callback code did not block in startcgocallback's approximation to exitsyscall, so if, say, the garbage collector were running, it would still barge in and start doing things like call malloc. Similarly endcgocallback's approximation of entersyscall did not call matchmg to kick off new OS threads when necessary, which caused the bug in issue 1560. Fixes #1560. R=iant CC=golang-dev https://golang.org/cl/4253054
2011-03-07 08:37:42 -07:00
// to keep running g.
TEXT runtime·mcall(SB), 7, $-4
MOVW fn+0(FP), R0
// Save caller state in g->sched.
runtime: scheduler, cgo reorganization * Change use of m->g0 stack (aka scheduler stack). * Provide runtime.mcall(f) to invoke f() on m->g0 stack. * Replace scheduler loop entry with runtime.mcall(schedule). Runtime.mcall eliminates the need for fake scheduler states that exist just to run a bit of code on the m->g0 stack (Grecovery, Gstackalloc). The elimination of the scheduler as a loop that stops and starts using gosave and gogo fixes a bad interaction with the way cgo uses the m->g0 stack. Cgo runs external (gcc-compiled) C functions on that stack, and then when calling back into Go, it sets m->g0->sched.sp below the added call frames, so that other uses of m->g0's stack will not interfere with those frames. Unfortunately, gogo (longjmp) back to the scheduler loop at this point would end up running scheduler with the lower sp, which no longer points at a valid stack frame for a call to scheduler. If scheduler then wrote any function call arguments or local variables to where it expected the stack frame to be, it would overwrite other data on the stack. I realized this possibility while debugging a problem with calling complex Go code in a Go -> C -> Go cgo callback. This wasn't the bug I was looking for, it turns out, but I believe it is a real bug nonetheless. Switching to runtime.mcall, which only adds new frames to the stack and never jumps into functions running in existing ones, fixes this bug. * Move cgo-related code out of proc.c into cgocall.c. * Add very large comment describing cgo call sequences. * Simpilify, regularize cgo function implementations and names. * Add test suite as misc/cgo/test. Now the Go -> C path calls cgocall, which calls asmcgocall, and the C -> Go path calls cgocallback, which calls cgocallbackg. The shuffling, which affects mainly the callback case, moves most of the callback implementation to cgocallback running on the m->curg stack (not the m->g0 scheduler stack) and only while accounted for with $GOMAXPROCS (between calls to exitsyscall and entersyscall). The previous callback code did not block in startcgocallback's approximation to exitsyscall, so if, say, the garbage collector were running, it would still barge in and start doing things like call malloc. Similarly endcgocallback's approximation of entersyscall did not call matchmg to kick off new OS threads when necessary, which caused the bug in issue 1560. Fixes #1560. R=iant CC=golang-dev https://golang.org/cl/4253054
2011-03-07 08:37:42 -07:00
MOVW SP, (g_sched+gobuf_sp)(g)
MOVW LR, (g_sched+gobuf_pc)(g)
runtime: record proper goroutine state during stack split Until now, the goroutine state has been scattered during the execution of newstack and oldstack. It's all there, and those routines know how to get back to a working goroutine, but other pieces of the system, like stack traces, do not. If something does interrupt the newstack or oldstack execution, the rest of the system can't understand the goroutine. For example, if newstack decides there is an overflow and calls throw, the stack tracer wouldn't dump the goroutine correctly. For newstack to save a useful state snapshot, it needs to be able to rewind the PC in the function that triggered the split back to the beginning of the function. (The PC is a few instructions in, just after the call to morestack.) To make that possible, we change the prologues to insert a jmp back to the beginning of the function after the call to morestack. That is, the prologue used to be roughly: TEXT myfunc check for split jmpcond nosplit call morestack nosplit: sub $xxx, sp Now an extra instruction is inserted after the call: TEXT myfunc start: check for split jmpcond nosplit call morestack jmp start nosplit: sub $xxx, sp The jmp is not executed directly. It is decoded and simulated by runtime.rewindmorestack to discover the beginning of the function, and then the call to morestack returns directly to the start label instead of to the jump instruction. So logically the jmp is still executed, just not by the cpu. The prologue thus repeats in the case of a function that needs a stack split, but against the cost of the split itself, the extra few instructions are noise. The repeated prologue has the nice effect of making a stack split double-check that the new stack is big enough: if morestack happens to return on a too-small stack, we'll now notice before corruption happens. The ability for newstack to rewind to the beginning of the function should help preemption too. If newstack decides that it was called for preemption instead of a stack split, it now has the goroutine state correctly paused if rescheduling is needed, and when the goroutine can run again, it can return to the start label on its original stack and re-execute the split check. Here is an example of a split stack overflow showing the full trace, without any special cases in the stack printer. (This one was triggered by making the split check incorrect.) runtime: newstack framesize=0x0 argsize=0x18 sp=0x6aebd0 stack=[0x6b0000, 0x6b0fa0] morebuf={pc:0x69f5b sp:0x6aebd8 lr:0x0} sched={pc:0x68880 sp:0x6aebd0 lr:0x0 ctxt:0x34e700} runtime: split stack overflow: 0x6aebd0 < 0x6b0000 fatal error: runtime: split stack overflow goroutine 1 [stack split]: runtime.mallocgc(0x290, 0x100000000, 0x1) /Users/rsc/g/go/src/pkg/runtime/zmalloc_darwin_amd64.c:21 fp=0x6aebd8 runtime.new() /Users/rsc/g/go/src/pkg/runtime/zmalloc_darwin_amd64.c:682 +0x5b fp=0x6aec08 go/build.(*Context).Import(0x5ae340, 0xc210030c71, 0xa, 0xc2100b4380, 0x1b, ...) /Users/rsc/g/go/src/pkg/go/build/build.go:424 +0x3a fp=0x6b00a0 main.loadImport(0xc210030c71, 0xa, 0xc2100b4380, 0x1b, 0xc2100b42c0, ...) /Users/rsc/g/go/src/cmd/go/pkg.go:249 +0x371 fp=0x6b01a8 main.(*Package).load(0xc21017c800, 0xc2100b42c0, 0xc2101828c0, 0x0, 0x0, ...) /Users/rsc/g/go/src/cmd/go/pkg.go:431 +0x2801 fp=0x6b0c98 main.loadPackage(0x369040, 0x7, 0xc2100b42c0, 0x0) /Users/rsc/g/go/src/cmd/go/pkg.go:709 +0x857 fp=0x6b0f80 ----- stack segment boundary ----- main.(*builder).action(0xc2100902a0, 0x0, 0x0, 0xc2100e6c00, 0xc2100e5750, ...) /Users/rsc/g/go/src/cmd/go/build.go:539 +0x437 fp=0x6b14a0 main.(*builder).action(0xc2100902a0, 0x0, 0x0, 0xc21015b400, 0x2, ...) /Users/rsc/g/go/src/cmd/go/build.go:528 +0x1d2 fp=0x6b1658 main.(*builder).test(0xc2100902a0, 0xc210092000, 0x0, 0x0, 0xc21008ff60, ...) /Users/rsc/g/go/src/cmd/go/test.go:622 +0x1b53 fp=0x6b1f68 ----- stack segment boundary ----- main.runTest(0x5a6b20, 0xc21000a020, 0x2, 0x2) /Users/rsc/g/go/src/cmd/go/test.go:366 +0xd09 fp=0x6a5cf0 main.main() /Users/rsc/g/go/src/cmd/go/main.go:161 +0x4f9 fp=0x6a5f78 runtime.main() /Users/rsc/g/go/src/pkg/runtime/proc.c:183 +0x92 fp=0x6a5fa0 runtime.goexit() /Users/rsc/g/go/src/pkg/runtime/proc.c:1266 fp=0x6a5fa8 And here is a seg fault during oldstack: SIGSEGV: segmentation violation PC=0x1b2a6 runtime.oldstack() /Users/rsc/g/go/src/pkg/runtime/stack.c:159 +0x76 runtime.lessstack() /Users/rsc/g/go/src/pkg/runtime/asm_amd64.s:270 +0x22 goroutine 1 [stack unsplit]: fmt.(*pp).printArg(0x2102e64e0, 0xe5c80, 0x2102c9220, 0x73, 0x0, ...) /Users/rsc/g/go/src/pkg/fmt/print.go:818 +0x3d3 fp=0x221031e6f8 fmt.(*pp).doPrintf(0x2102e64e0, 0x12fb20, 0x2, 0x221031eb98, 0x1, ...) /Users/rsc/g/go/src/pkg/fmt/print.go:1183 +0x15cb fp=0x221031eaf0 fmt.Sprintf(0x12fb20, 0x2, 0x221031eb98, 0x1, 0x1, ...) /Users/rsc/g/go/src/pkg/fmt/print.go:234 +0x67 fp=0x221031eb40 flag.(*stringValue).String(0x2102c9210, 0x1, 0x0) /Users/rsc/g/go/src/pkg/flag/flag.go:180 +0xb3 fp=0x221031ebb0 flag.(*FlagSet).Var(0x2102f6000, 0x293d38, 0x2102c9210, 0x143490, 0xa, ...) /Users/rsc/g/go/src/pkg/flag/flag.go:633 +0x40 fp=0x221031eca0 flag.(*FlagSet).StringVar(0x2102f6000, 0x2102c9210, 0x143490, 0xa, 0x12fa60, ...) /Users/rsc/g/go/src/pkg/flag/flag.go:550 +0x91 fp=0x221031ece8 flag.(*FlagSet).String(0x2102f6000, 0x143490, 0xa, 0x12fa60, 0x0, ...) /Users/rsc/g/go/src/pkg/flag/flag.go:563 +0x87 fp=0x221031ed38 flag.String(0x143490, 0xa, 0x12fa60, 0x0, 0x161950, ...) /Users/rsc/g/go/src/pkg/flag/flag.go:570 +0x6b fp=0x221031ed80 testing.init() /Users/rsc/g/go/src/pkg/testing/testing.go:-531 +0xbb fp=0x221031edc0 strings_test.init() /Users/rsc/g/go/src/pkg/strings/strings_test.go:1115 +0x62 fp=0x221031ef70 main.init() strings/_test/_testmain.go:90 +0x3d fp=0x221031ef78 runtime.main() /Users/rsc/g/go/src/pkg/runtime/proc.c:180 +0x8a fp=0x221031efa0 runtime.goexit() /Users/rsc/g/go/src/pkg/runtime/proc.c:1269 fp=0x221031efa8 goroutine 2 [runnable]: runtime.MHeap_Scavenger() /Users/rsc/g/go/src/pkg/runtime/mheap.c:438 runtime.goexit() /Users/rsc/g/go/src/pkg/runtime/proc.c:1269 created by runtime.main /Users/rsc/g/go/src/pkg/runtime/proc.c:166 rax 0x23ccc0 rbx 0x23ccc0 rcx 0x0 rdx 0x38 rdi 0x2102c0170 rsi 0x221032cfe0 rbp 0x221032cfa0 rsp 0x7fff5fbff5b0 r8 0x2102c0120 r9 0x221032cfa0 r10 0x221032c000 r11 0x104ce8 r12 0xe5c80 r13 0x1be82baac718 r14 0x13091135f7d69200 r15 0x0 rip 0x1b2a6 rflags 0x10246 cs 0x2b fs 0x0 gs 0x0 Fixes #5723. R=r, dvyukov, go.peter.90, dave, iant CC=golang-dev https://golang.org/cl/10360048
2013-06-27 09:32:01 -06:00
MOVW $0, R11
MOVW R11, (g_sched+gobuf_lr)(g)
runtime: scheduler, cgo reorganization * Change use of m->g0 stack (aka scheduler stack). * Provide runtime.mcall(f) to invoke f() on m->g0 stack. * Replace scheduler loop entry with runtime.mcall(schedule). Runtime.mcall eliminates the need for fake scheduler states that exist just to run a bit of code on the m->g0 stack (Grecovery, Gstackalloc). The elimination of the scheduler as a loop that stops and starts using gosave and gogo fixes a bad interaction with the way cgo uses the m->g0 stack. Cgo runs external (gcc-compiled) C functions on that stack, and then when calling back into Go, it sets m->g0->sched.sp below the added call frames, so that other uses of m->g0's stack will not interfere with those frames. Unfortunately, gogo (longjmp) back to the scheduler loop at this point would end up running scheduler with the lower sp, which no longer points at a valid stack frame for a call to scheduler. If scheduler then wrote any function call arguments or local variables to where it expected the stack frame to be, it would overwrite other data on the stack. I realized this possibility while debugging a problem with calling complex Go code in a Go -> C -> Go cgo callback. This wasn't the bug I was looking for, it turns out, but I believe it is a real bug nonetheless. Switching to runtime.mcall, which only adds new frames to the stack and never jumps into functions running in existing ones, fixes this bug. * Move cgo-related code out of proc.c into cgocall.c. * Add very large comment describing cgo call sequences. * Simpilify, regularize cgo function implementations and names. * Add test suite as misc/cgo/test. Now the Go -> C path calls cgocall, which calls asmcgocall, and the C -> Go path calls cgocallback, which calls cgocallbackg. The shuffling, which affects mainly the callback case, moves most of the callback implementation to cgocallback running on the m->curg stack (not the m->g0 scheduler stack) and only while accounted for with $GOMAXPROCS (between calls to exitsyscall and entersyscall). The previous callback code did not block in startcgocallback's approximation to exitsyscall, so if, say, the garbage collector were running, it would still barge in and start doing things like call malloc. Similarly endcgocallback's approximation of entersyscall did not call matchmg to kick off new OS threads when necessary, which caused the bug in issue 1560. Fixes #1560. R=iant CC=golang-dev https://golang.org/cl/4253054
2011-03-07 08:37:42 -07:00
MOVW g, (g_sched+gobuf_g)(g)
// Switch to m->g0 & its stack, call fn.
MOVW g, R1
MOVW m_g0(m), g
CMP g, R1
BL.EQ runtime·badmcall(SB)
MOVW (g_sched+gobuf_sp)(g), SP
SUB $8, SP
MOVW R1, 4(SP)
BL (R0)
BL runtime·badmcall2(SB)
RET
/*
* support for morestack
*/
// Called during function prolog when more stack is needed.
// R1 frame size
// R2 arg size
// R3 prolog's LR
// NB. we do not save R0 because we've forced 5c to pass all arguments
// on the stack.
// using frame size $-4 means do not save LR on stack.
TEXT runtime·morestack(SB),7,$-4
// Cannot grow scheduler stack (m->g0).
MOVW m_g0(m), R4
CMP g, R4
BL.EQ runtime·abort(SB)
MOVW R1, m_moreframesize(m)
MOVW R2, m_moreargsize(m)
runtime: record proper goroutine state during stack split Until now, the goroutine state has been scattered during the execution of newstack and oldstack. It's all there, and those routines know how to get back to a working goroutine, but other pieces of the system, like stack traces, do not. If something does interrupt the newstack or oldstack execution, the rest of the system can't understand the goroutine. For example, if newstack decides there is an overflow and calls throw, the stack tracer wouldn't dump the goroutine correctly. For newstack to save a useful state snapshot, it needs to be able to rewind the PC in the function that triggered the split back to the beginning of the function. (The PC is a few instructions in, just after the call to morestack.) To make that possible, we change the prologues to insert a jmp back to the beginning of the function after the call to morestack. That is, the prologue used to be roughly: TEXT myfunc check for split jmpcond nosplit call morestack nosplit: sub $xxx, sp Now an extra instruction is inserted after the call: TEXT myfunc start: check for split jmpcond nosplit call morestack jmp start nosplit: sub $xxx, sp The jmp is not executed directly. It is decoded and simulated by runtime.rewindmorestack to discover the beginning of the function, and then the call to morestack returns directly to the start label instead of to the jump instruction. So logically the jmp is still executed, just not by the cpu. The prologue thus repeats in the case of a function that needs a stack split, but against the cost of the split itself, the extra few instructions are noise. The repeated prologue has the nice effect of making a stack split double-check that the new stack is big enough: if morestack happens to return on a too-small stack, we'll now notice before corruption happens. The ability for newstack to rewind to the beginning of the function should help preemption too. If newstack decides that it was called for preemption instead of a stack split, it now has the goroutine state correctly paused if rescheduling is needed, and when the goroutine can run again, it can return to the start label on its original stack and re-execute the split check. Here is an example of a split stack overflow showing the full trace, without any special cases in the stack printer. (This one was triggered by making the split check incorrect.) runtime: newstack framesize=0x0 argsize=0x18 sp=0x6aebd0 stack=[0x6b0000, 0x6b0fa0] morebuf={pc:0x69f5b sp:0x6aebd8 lr:0x0} sched={pc:0x68880 sp:0x6aebd0 lr:0x0 ctxt:0x34e700} runtime: split stack overflow: 0x6aebd0 < 0x6b0000 fatal error: runtime: split stack overflow goroutine 1 [stack split]: runtime.mallocgc(0x290, 0x100000000, 0x1) /Users/rsc/g/go/src/pkg/runtime/zmalloc_darwin_amd64.c:21 fp=0x6aebd8 runtime.new() /Users/rsc/g/go/src/pkg/runtime/zmalloc_darwin_amd64.c:682 +0x5b fp=0x6aec08 go/build.(*Context).Import(0x5ae340, 0xc210030c71, 0xa, 0xc2100b4380, 0x1b, ...) /Users/rsc/g/go/src/pkg/go/build/build.go:424 +0x3a fp=0x6b00a0 main.loadImport(0xc210030c71, 0xa, 0xc2100b4380, 0x1b, 0xc2100b42c0, ...) /Users/rsc/g/go/src/cmd/go/pkg.go:249 +0x371 fp=0x6b01a8 main.(*Package).load(0xc21017c800, 0xc2100b42c0, 0xc2101828c0, 0x0, 0x0, ...) /Users/rsc/g/go/src/cmd/go/pkg.go:431 +0x2801 fp=0x6b0c98 main.loadPackage(0x369040, 0x7, 0xc2100b42c0, 0x0) /Users/rsc/g/go/src/cmd/go/pkg.go:709 +0x857 fp=0x6b0f80 ----- stack segment boundary ----- main.(*builder).action(0xc2100902a0, 0x0, 0x0, 0xc2100e6c00, 0xc2100e5750, ...) /Users/rsc/g/go/src/cmd/go/build.go:539 +0x437 fp=0x6b14a0 main.(*builder).action(0xc2100902a0, 0x0, 0x0, 0xc21015b400, 0x2, ...) /Users/rsc/g/go/src/cmd/go/build.go:528 +0x1d2 fp=0x6b1658 main.(*builder).test(0xc2100902a0, 0xc210092000, 0x0, 0x0, 0xc21008ff60, ...) /Users/rsc/g/go/src/cmd/go/test.go:622 +0x1b53 fp=0x6b1f68 ----- stack segment boundary ----- main.runTest(0x5a6b20, 0xc21000a020, 0x2, 0x2) /Users/rsc/g/go/src/cmd/go/test.go:366 +0xd09 fp=0x6a5cf0 main.main() /Users/rsc/g/go/src/cmd/go/main.go:161 +0x4f9 fp=0x6a5f78 runtime.main() /Users/rsc/g/go/src/pkg/runtime/proc.c:183 +0x92 fp=0x6a5fa0 runtime.goexit() /Users/rsc/g/go/src/pkg/runtime/proc.c:1266 fp=0x6a5fa8 And here is a seg fault during oldstack: SIGSEGV: segmentation violation PC=0x1b2a6 runtime.oldstack() /Users/rsc/g/go/src/pkg/runtime/stack.c:159 +0x76 runtime.lessstack() /Users/rsc/g/go/src/pkg/runtime/asm_amd64.s:270 +0x22 goroutine 1 [stack unsplit]: fmt.(*pp).printArg(0x2102e64e0, 0xe5c80, 0x2102c9220, 0x73, 0x0, ...) /Users/rsc/g/go/src/pkg/fmt/print.go:818 +0x3d3 fp=0x221031e6f8 fmt.(*pp).doPrintf(0x2102e64e0, 0x12fb20, 0x2, 0x221031eb98, 0x1, ...) /Users/rsc/g/go/src/pkg/fmt/print.go:1183 +0x15cb fp=0x221031eaf0 fmt.Sprintf(0x12fb20, 0x2, 0x221031eb98, 0x1, 0x1, ...) /Users/rsc/g/go/src/pkg/fmt/print.go:234 +0x67 fp=0x221031eb40 flag.(*stringValue).String(0x2102c9210, 0x1, 0x0) /Users/rsc/g/go/src/pkg/flag/flag.go:180 +0xb3 fp=0x221031ebb0 flag.(*FlagSet).Var(0x2102f6000, 0x293d38, 0x2102c9210, 0x143490, 0xa, ...) /Users/rsc/g/go/src/pkg/flag/flag.go:633 +0x40 fp=0x221031eca0 flag.(*FlagSet).StringVar(0x2102f6000, 0x2102c9210, 0x143490, 0xa, 0x12fa60, ...) /Users/rsc/g/go/src/pkg/flag/flag.go:550 +0x91 fp=0x221031ece8 flag.(*FlagSet).String(0x2102f6000, 0x143490, 0xa, 0x12fa60, 0x0, ...) /Users/rsc/g/go/src/pkg/flag/flag.go:563 +0x87 fp=0x221031ed38 flag.String(0x143490, 0xa, 0x12fa60, 0x0, 0x161950, ...) /Users/rsc/g/go/src/pkg/flag/flag.go:570 +0x6b fp=0x221031ed80 testing.init() /Users/rsc/g/go/src/pkg/testing/testing.go:-531 +0xbb fp=0x221031edc0 strings_test.init() /Users/rsc/g/go/src/pkg/strings/strings_test.go:1115 +0x62 fp=0x221031ef70 main.init() strings/_test/_testmain.go:90 +0x3d fp=0x221031ef78 runtime.main() /Users/rsc/g/go/src/pkg/runtime/proc.c:180 +0x8a fp=0x221031efa0 runtime.goexit() /Users/rsc/g/go/src/pkg/runtime/proc.c:1269 fp=0x221031efa8 goroutine 2 [runnable]: runtime.MHeap_Scavenger() /Users/rsc/g/go/src/pkg/runtime/mheap.c:438 runtime.goexit() /Users/rsc/g/go/src/pkg/runtime/proc.c:1269 created by runtime.main /Users/rsc/g/go/src/pkg/runtime/proc.c:166 rax 0x23ccc0 rbx 0x23ccc0 rcx 0x0 rdx 0x38 rdi 0x2102c0170 rsi 0x221032cfe0 rbp 0x221032cfa0 rsp 0x7fff5fbff5b0 r8 0x2102c0120 r9 0x221032cfa0 r10 0x221032c000 r11 0x104ce8 r12 0xe5c80 r13 0x1be82baac718 r14 0x13091135f7d69200 r15 0x0 rip 0x1b2a6 rflags 0x10246 cs 0x2b fs 0x0 gs 0x0 Fixes #5723. R=r, dvyukov, go.peter.90, dave, iant CC=golang-dev https://golang.org/cl/10360048
2013-06-27 09:32:01 -06:00
// Called from f.
// Set g->sched to context in f.
MOVW R7, (g_sched+gobuf_ctxt)(g)
MOVW SP, (g_sched+gobuf_sp)(g)
MOVW LR, (g_sched+gobuf_pc)(g)
MOVW R3, (g_sched+gobuf_lr)(g)
// Called from f.
// Set m->morebuf to f's caller.
MOVW R3, (m_morebuf+gobuf_pc)(m) // f's caller's PC
MOVW SP, (m_morebuf+gobuf_sp)(m) // f's caller's SP
MOVW $4(SP), R3 // f's argument pointer
MOVW R3, m_moreargp(m)
MOVW g, (m_morebuf+gobuf_g)(m)
runtime: scheduler, cgo reorganization * Change use of m->g0 stack (aka scheduler stack). * Provide runtime.mcall(f) to invoke f() on m->g0 stack. * Replace scheduler loop entry with runtime.mcall(schedule). Runtime.mcall eliminates the need for fake scheduler states that exist just to run a bit of code on the m->g0 stack (Grecovery, Gstackalloc). The elimination of the scheduler as a loop that stops and starts using gosave and gogo fixes a bad interaction with the way cgo uses the m->g0 stack. Cgo runs external (gcc-compiled) C functions on that stack, and then when calling back into Go, it sets m->g0->sched.sp below the added call frames, so that other uses of m->g0's stack will not interfere with those frames. Unfortunately, gogo (longjmp) back to the scheduler loop at this point would end up running scheduler with the lower sp, which no longer points at a valid stack frame for a call to scheduler. If scheduler then wrote any function call arguments or local variables to where it expected the stack frame to be, it would overwrite other data on the stack. I realized this possibility while debugging a problem with calling complex Go code in a Go -> C -> Go cgo callback. This wasn't the bug I was looking for, it turns out, but I believe it is a real bug nonetheless. Switching to runtime.mcall, which only adds new frames to the stack and never jumps into functions running in existing ones, fixes this bug. * Move cgo-related code out of proc.c into cgocall.c. * Add very large comment describing cgo call sequences. * Simpilify, regularize cgo function implementations and names. * Add test suite as misc/cgo/test. Now the Go -> C path calls cgocall, which calls asmcgocall, and the C -> Go path calls cgocallback, which calls cgocallbackg. The shuffling, which affects mainly the callback case, moves most of the callback implementation to cgocallback running on the m->curg stack (not the m->g0 scheduler stack) and only while accounted for with $GOMAXPROCS (between calls to exitsyscall and entersyscall). The previous callback code did not block in startcgocallback's approximation to exitsyscall, so if, say, the garbage collector were running, it would still barge in and start doing things like call malloc. Similarly endcgocallback's approximation of entersyscall did not call matchmg to kick off new OS threads when necessary, which caused the bug in issue 1560. Fixes #1560. R=iant CC=golang-dev https://golang.org/cl/4253054
2011-03-07 08:37:42 -07:00
// Call newstack on m->g0's stack.
MOVW m_g0(m), g
runtime: scheduler, cgo reorganization * Change use of m->g0 stack (aka scheduler stack). * Provide runtime.mcall(f) to invoke f() on m->g0 stack. * Replace scheduler loop entry with runtime.mcall(schedule). Runtime.mcall eliminates the need for fake scheduler states that exist just to run a bit of code on the m->g0 stack (Grecovery, Gstackalloc). The elimination of the scheduler as a loop that stops and starts using gosave and gogo fixes a bad interaction with the way cgo uses the m->g0 stack. Cgo runs external (gcc-compiled) C functions on that stack, and then when calling back into Go, it sets m->g0->sched.sp below the added call frames, so that other uses of m->g0's stack will not interfere with those frames. Unfortunately, gogo (longjmp) back to the scheduler loop at this point would end up running scheduler with the lower sp, which no longer points at a valid stack frame for a call to scheduler. If scheduler then wrote any function call arguments or local variables to where it expected the stack frame to be, it would overwrite other data on the stack. I realized this possibility while debugging a problem with calling complex Go code in a Go -> C -> Go cgo callback. This wasn't the bug I was looking for, it turns out, but I believe it is a real bug nonetheless. Switching to runtime.mcall, which only adds new frames to the stack and never jumps into functions running in existing ones, fixes this bug. * Move cgo-related code out of proc.c into cgocall.c. * Add very large comment describing cgo call sequences. * Simpilify, regularize cgo function implementations and names. * Add test suite as misc/cgo/test. Now the Go -> C path calls cgocall, which calls asmcgocall, and the C -> Go path calls cgocallback, which calls cgocallbackg. The shuffling, which affects mainly the callback case, moves most of the callback implementation to cgocallback running on the m->curg stack (not the m->g0 scheduler stack) and only while accounted for with $GOMAXPROCS (between calls to exitsyscall and entersyscall). The previous callback code did not block in startcgocallback's approximation to exitsyscall, so if, say, the garbage collector were running, it would still barge in and start doing things like call malloc. Similarly endcgocallback's approximation of entersyscall did not call matchmg to kick off new OS threads when necessary, which caused the bug in issue 1560. Fixes #1560. R=iant CC=golang-dev https://golang.org/cl/4253054
2011-03-07 08:37:42 -07:00
MOVW (g_sched+gobuf_sp)(g), SP
B runtime·newstack(SB)
// Called from reflection library. Mimics morestack,
// reuses stack growth code to create a frame
// with the desired args running the desired function.
//
// func call(fn *byte, arg *byte, argsize uint32).
TEXT reflect·call(SB), 7, $-4
// Save our caller's state as the PC and SP to
// restore when returning from f.
MOVW LR, (m_morebuf+gobuf_pc)(m) // our caller's PC
MOVW SP, (m_morebuf+gobuf_sp)(m) // our caller's SP
MOVW g, (m_morebuf+gobuf_g)(m)
// Save our own state as the PC and SP to restore
// if this goroutine needs to be restarted.
MOVW $reflect·call(SB), R11
MOVW R11, (g_sched+gobuf_pc)(g)
MOVW LR, (g_sched+gobuf_lr)(g)
MOVW SP, (g_sched+gobuf_sp)(g)
// Set up morestack arguments to call f on a new stack.
// We set f's frame size to 1, as a hint to newstack
// that this is a call from reflect·call.
// If it turns out that f needs a larger frame than
// the default stack, f's usual stack growth prolog will
// allocate a new segment (and recopy the arguments).
MOVW 4(SP), R0 // fn
MOVW 8(SP), R1 // arg frame
MOVW 12(SP), R2 // arg size
runtime: record proper goroutine state during stack split Until now, the goroutine state has been scattered during the execution of newstack and oldstack. It's all there, and those routines know how to get back to a working goroutine, but other pieces of the system, like stack traces, do not. If something does interrupt the newstack or oldstack execution, the rest of the system can't understand the goroutine. For example, if newstack decides there is an overflow and calls throw, the stack tracer wouldn't dump the goroutine correctly. For newstack to save a useful state snapshot, it needs to be able to rewind the PC in the function that triggered the split back to the beginning of the function. (The PC is a few instructions in, just after the call to morestack.) To make that possible, we change the prologues to insert a jmp back to the beginning of the function after the call to morestack. That is, the prologue used to be roughly: TEXT myfunc check for split jmpcond nosplit call morestack nosplit: sub $xxx, sp Now an extra instruction is inserted after the call: TEXT myfunc start: check for split jmpcond nosplit call morestack jmp start nosplit: sub $xxx, sp The jmp is not executed directly. It is decoded and simulated by runtime.rewindmorestack to discover the beginning of the function, and then the call to morestack returns directly to the start label instead of to the jump instruction. So logically the jmp is still executed, just not by the cpu. The prologue thus repeats in the case of a function that needs a stack split, but against the cost of the split itself, the extra few instructions are noise. The repeated prologue has the nice effect of making a stack split double-check that the new stack is big enough: if morestack happens to return on a too-small stack, we'll now notice before corruption happens. The ability for newstack to rewind to the beginning of the function should help preemption too. If newstack decides that it was called for preemption instead of a stack split, it now has the goroutine state correctly paused if rescheduling is needed, and when the goroutine can run again, it can return to the start label on its original stack and re-execute the split check. Here is an example of a split stack overflow showing the full trace, without any special cases in the stack printer. (This one was triggered by making the split check incorrect.) runtime: newstack framesize=0x0 argsize=0x18 sp=0x6aebd0 stack=[0x6b0000, 0x6b0fa0] morebuf={pc:0x69f5b sp:0x6aebd8 lr:0x0} sched={pc:0x68880 sp:0x6aebd0 lr:0x0 ctxt:0x34e700} runtime: split stack overflow: 0x6aebd0 < 0x6b0000 fatal error: runtime: split stack overflow goroutine 1 [stack split]: runtime.mallocgc(0x290, 0x100000000, 0x1) /Users/rsc/g/go/src/pkg/runtime/zmalloc_darwin_amd64.c:21 fp=0x6aebd8 runtime.new() /Users/rsc/g/go/src/pkg/runtime/zmalloc_darwin_amd64.c:682 +0x5b fp=0x6aec08 go/build.(*Context).Import(0x5ae340, 0xc210030c71, 0xa, 0xc2100b4380, 0x1b, ...) /Users/rsc/g/go/src/pkg/go/build/build.go:424 +0x3a fp=0x6b00a0 main.loadImport(0xc210030c71, 0xa, 0xc2100b4380, 0x1b, 0xc2100b42c0, ...) /Users/rsc/g/go/src/cmd/go/pkg.go:249 +0x371 fp=0x6b01a8 main.(*Package).load(0xc21017c800, 0xc2100b42c0, 0xc2101828c0, 0x0, 0x0, ...) /Users/rsc/g/go/src/cmd/go/pkg.go:431 +0x2801 fp=0x6b0c98 main.loadPackage(0x369040, 0x7, 0xc2100b42c0, 0x0) /Users/rsc/g/go/src/cmd/go/pkg.go:709 +0x857 fp=0x6b0f80 ----- stack segment boundary ----- main.(*builder).action(0xc2100902a0, 0x0, 0x0, 0xc2100e6c00, 0xc2100e5750, ...) /Users/rsc/g/go/src/cmd/go/build.go:539 +0x437 fp=0x6b14a0 main.(*builder).action(0xc2100902a0, 0x0, 0x0, 0xc21015b400, 0x2, ...) /Users/rsc/g/go/src/cmd/go/build.go:528 +0x1d2 fp=0x6b1658 main.(*builder).test(0xc2100902a0, 0xc210092000, 0x0, 0x0, 0xc21008ff60, ...) /Users/rsc/g/go/src/cmd/go/test.go:622 +0x1b53 fp=0x6b1f68 ----- stack segment boundary ----- main.runTest(0x5a6b20, 0xc21000a020, 0x2, 0x2) /Users/rsc/g/go/src/cmd/go/test.go:366 +0xd09 fp=0x6a5cf0 main.main() /Users/rsc/g/go/src/cmd/go/main.go:161 +0x4f9 fp=0x6a5f78 runtime.main() /Users/rsc/g/go/src/pkg/runtime/proc.c:183 +0x92 fp=0x6a5fa0 runtime.goexit() /Users/rsc/g/go/src/pkg/runtime/proc.c:1266 fp=0x6a5fa8 And here is a seg fault during oldstack: SIGSEGV: segmentation violation PC=0x1b2a6 runtime.oldstack() /Users/rsc/g/go/src/pkg/runtime/stack.c:159 +0x76 runtime.lessstack() /Users/rsc/g/go/src/pkg/runtime/asm_amd64.s:270 +0x22 goroutine 1 [stack unsplit]: fmt.(*pp).printArg(0x2102e64e0, 0xe5c80, 0x2102c9220, 0x73, 0x0, ...) /Users/rsc/g/go/src/pkg/fmt/print.go:818 +0x3d3 fp=0x221031e6f8 fmt.(*pp).doPrintf(0x2102e64e0, 0x12fb20, 0x2, 0x221031eb98, 0x1, ...) /Users/rsc/g/go/src/pkg/fmt/print.go:1183 +0x15cb fp=0x221031eaf0 fmt.Sprintf(0x12fb20, 0x2, 0x221031eb98, 0x1, 0x1, ...) /Users/rsc/g/go/src/pkg/fmt/print.go:234 +0x67 fp=0x221031eb40 flag.(*stringValue).String(0x2102c9210, 0x1, 0x0) /Users/rsc/g/go/src/pkg/flag/flag.go:180 +0xb3 fp=0x221031ebb0 flag.(*FlagSet).Var(0x2102f6000, 0x293d38, 0x2102c9210, 0x143490, 0xa, ...) /Users/rsc/g/go/src/pkg/flag/flag.go:633 +0x40 fp=0x221031eca0 flag.(*FlagSet).StringVar(0x2102f6000, 0x2102c9210, 0x143490, 0xa, 0x12fa60, ...) /Users/rsc/g/go/src/pkg/flag/flag.go:550 +0x91 fp=0x221031ece8 flag.(*FlagSet).String(0x2102f6000, 0x143490, 0xa, 0x12fa60, 0x0, ...) /Users/rsc/g/go/src/pkg/flag/flag.go:563 +0x87 fp=0x221031ed38 flag.String(0x143490, 0xa, 0x12fa60, 0x0, 0x161950, ...) /Users/rsc/g/go/src/pkg/flag/flag.go:570 +0x6b fp=0x221031ed80 testing.init() /Users/rsc/g/go/src/pkg/testing/testing.go:-531 +0xbb fp=0x221031edc0 strings_test.init() /Users/rsc/g/go/src/pkg/strings/strings_test.go:1115 +0x62 fp=0x221031ef70 main.init() strings/_test/_testmain.go:90 +0x3d fp=0x221031ef78 runtime.main() /Users/rsc/g/go/src/pkg/runtime/proc.c:180 +0x8a fp=0x221031efa0 runtime.goexit() /Users/rsc/g/go/src/pkg/runtime/proc.c:1269 fp=0x221031efa8 goroutine 2 [runnable]: runtime.MHeap_Scavenger() /Users/rsc/g/go/src/pkg/runtime/mheap.c:438 runtime.goexit() /Users/rsc/g/go/src/pkg/runtime/proc.c:1269 created by runtime.main /Users/rsc/g/go/src/pkg/runtime/proc.c:166 rax 0x23ccc0 rbx 0x23ccc0 rcx 0x0 rdx 0x38 rdi 0x2102c0170 rsi 0x221032cfe0 rbp 0x221032cfa0 rsp 0x7fff5fbff5b0 r8 0x2102c0120 r9 0x221032cfa0 r10 0x221032c000 r11 0x104ce8 r12 0xe5c80 r13 0x1be82baac718 r14 0x13091135f7d69200 r15 0x0 rip 0x1b2a6 rflags 0x10246 cs 0x2b fs 0x0 gs 0x0 Fixes #5723. R=r, dvyukov, go.peter.90, dave, iant CC=golang-dev https://golang.org/cl/10360048
2013-06-27 09:32:01 -06:00
MOVW R0, m_cret(m) // f's PC
MOVW R1, m_moreargp(m) // f's argument pointer
MOVW R2, m_moreargsize(m) // f's argument size
MOVW $1, R3
MOVW R3, m_moreframesize(m) // f's frame size
runtime: scheduler, cgo reorganization * Change use of m->g0 stack (aka scheduler stack). * Provide runtime.mcall(f) to invoke f() on m->g0 stack. * Replace scheduler loop entry with runtime.mcall(schedule). Runtime.mcall eliminates the need for fake scheduler states that exist just to run a bit of code on the m->g0 stack (Grecovery, Gstackalloc). The elimination of the scheduler as a loop that stops and starts using gosave and gogo fixes a bad interaction with the way cgo uses the m->g0 stack. Cgo runs external (gcc-compiled) C functions on that stack, and then when calling back into Go, it sets m->g0->sched.sp below the added call frames, so that other uses of m->g0's stack will not interfere with those frames. Unfortunately, gogo (longjmp) back to the scheduler loop at this point would end up running scheduler with the lower sp, which no longer points at a valid stack frame for a call to scheduler. If scheduler then wrote any function call arguments or local variables to where it expected the stack frame to be, it would overwrite other data on the stack. I realized this possibility while debugging a problem with calling complex Go code in a Go -> C -> Go cgo callback. This wasn't the bug I was looking for, it turns out, but I believe it is a real bug nonetheless. Switching to runtime.mcall, which only adds new frames to the stack and never jumps into functions running in existing ones, fixes this bug. * Move cgo-related code out of proc.c into cgocall.c. * Add very large comment describing cgo call sequences. * Simpilify, regularize cgo function implementations and names. * Add test suite as misc/cgo/test. Now the Go -> C path calls cgocall, which calls asmcgocall, and the C -> Go path calls cgocallback, which calls cgocallbackg. The shuffling, which affects mainly the callback case, moves most of the callback implementation to cgocallback running on the m->curg stack (not the m->g0 scheduler stack) and only while accounted for with $GOMAXPROCS (between calls to exitsyscall and entersyscall). The previous callback code did not block in startcgocallback's approximation to exitsyscall, so if, say, the garbage collector were running, it would still barge in and start doing things like call malloc. Similarly endcgocallback's approximation of entersyscall did not call matchmg to kick off new OS threads when necessary, which caused the bug in issue 1560. Fixes #1560. R=iant CC=golang-dev https://golang.org/cl/4253054
2011-03-07 08:37:42 -07:00
// Call newstack on m->g0's stack.
MOVW m_g0(m), g
runtime: scheduler, cgo reorganization * Change use of m->g0 stack (aka scheduler stack). * Provide runtime.mcall(f) to invoke f() on m->g0 stack. * Replace scheduler loop entry with runtime.mcall(schedule). Runtime.mcall eliminates the need for fake scheduler states that exist just to run a bit of code on the m->g0 stack (Grecovery, Gstackalloc). The elimination of the scheduler as a loop that stops and starts using gosave and gogo fixes a bad interaction with the way cgo uses the m->g0 stack. Cgo runs external (gcc-compiled) C functions on that stack, and then when calling back into Go, it sets m->g0->sched.sp below the added call frames, so that other uses of m->g0's stack will not interfere with those frames. Unfortunately, gogo (longjmp) back to the scheduler loop at this point would end up running scheduler with the lower sp, which no longer points at a valid stack frame for a call to scheduler. If scheduler then wrote any function call arguments or local variables to where it expected the stack frame to be, it would overwrite other data on the stack. I realized this possibility while debugging a problem with calling complex Go code in a Go -> C -> Go cgo callback. This wasn't the bug I was looking for, it turns out, but I believe it is a real bug nonetheless. Switching to runtime.mcall, which only adds new frames to the stack and never jumps into functions running in existing ones, fixes this bug. * Move cgo-related code out of proc.c into cgocall.c. * Add very large comment describing cgo call sequences. * Simpilify, regularize cgo function implementations and names. * Add test suite as misc/cgo/test. Now the Go -> C path calls cgocall, which calls asmcgocall, and the C -> Go path calls cgocallback, which calls cgocallbackg. The shuffling, which affects mainly the callback case, moves most of the callback implementation to cgocallback running on the m->curg stack (not the m->g0 scheduler stack) and only while accounted for with $GOMAXPROCS (between calls to exitsyscall and entersyscall). The previous callback code did not block in startcgocallback's approximation to exitsyscall, so if, say, the garbage collector were running, it would still barge in and start doing things like call malloc. Similarly endcgocallback's approximation of entersyscall did not call matchmg to kick off new OS threads when necessary, which caused the bug in issue 1560. Fixes #1560. R=iant CC=golang-dev https://golang.org/cl/4253054
2011-03-07 08:37:42 -07:00
MOVW (g_sched+gobuf_sp)(g), SP
B runtime·newstack(SB)
// Return point when leaving stack.
// using frame size $-4 means do not save LR on stack.
TEXT runtime·lessstack(SB), 7, $-4
// Save return value in m->cret
MOVW R0, m_cret(m)
runtime: scheduler, cgo reorganization * Change use of m->g0 stack (aka scheduler stack). * Provide runtime.mcall(f) to invoke f() on m->g0 stack. * Replace scheduler loop entry with runtime.mcall(schedule). Runtime.mcall eliminates the need for fake scheduler states that exist just to run a bit of code on the m->g0 stack (Grecovery, Gstackalloc). The elimination of the scheduler as a loop that stops and starts using gosave and gogo fixes a bad interaction with the way cgo uses the m->g0 stack. Cgo runs external (gcc-compiled) C functions on that stack, and then when calling back into Go, it sets m->g0->sched.sp below the added call frames, so that other uses of m->g0's stack will not interfere with those frames. Unfortunately, gogo (longjmp) back to the scheduler loop at this point would end up running scheduler with the lower sp, which no longer points at a valid stack frame for a call to scheduler. If scheduler then wrote any function call arguments or local variables to where it expected the stack frame to be, it would overwrite other data on the stack. I realized this possibility while debugging a problem with calling complex Go code in a Go -> C -> Go cgo callback. This wasn't the bug I was looking for, it turns out, but I believe it is a real bug nonetheless. Switching to runtime.mcall, which only adds new frames to the stack and never jumps into functions running in existing ones, fixes this bug. * Move cgo-related code out of proc.c into cgocall.c. * Add very large comment describing cgo call sequences. * Simpilify, regularize cgo function implementations and names. * Add test suite as misc/cgo/test. Now the Go -> C path calls cgocall, which calls asmcgocall, and the C -> Go path calls cgocallback, which calls cgocallbackg. The shuffling, which affects mainly the callback case, moves most of the callback implementation to cgocallback running on the m->curg stack (not the m->g0 scheduler stack) and only while accounted for with $GOMAXPROCS (between calls to exitsyscall and entersyscall). The previous callback code did not block in startcgocallback's approximation to exitsyscall, so if, say, the garbage collector were running, it would still barge in and start doing things like call malloc. Similarly endcgocallback's approximation of entersyscall did not call matchmg to kick off new OS threads when necessary, which caused the bug in issue 1560. Fixes #1560. R=iant CC=golang-dev https://golang.org/cl/4253054
2011-03-07 08:37:42 -07:00
// Call oldstack on m->g0's stack.
MOVW m_g0(m), g
runtime: scheduler, cgo reorganization * Change use of m->g0 stack (aka scheduler stack). * Provide runtime.mcall(f) to invoke f() on m->g0 stack. * Replace scheduler loop entry with runtime.mcall(schedule). Runtime.mcall eliminates the need for fake scheduler states that exist just to run a bit of code on the m->g0 stack (Grecovery, Gstackalloc). The elimination of the scheduler as a loop that stops and starts using gosave and gogo fixes a bad interaction with the way cgo uses the m->g0 stack. Cgo runs external (gcc-compiled) C functions on that stack, and then when calling back into Go, it sets m->g0->sched.sp below the added call frames, so that other uses of m->g0's stack will not interfere with those frames. Unfortunately, gogo (longjmp) back to the scheduler loop at this point would end up running scheduler with the lower sp, which no longer points at a valid stack frame for a call to scheduler. If scheduler then wrote any function call arguments or local variables to where it expected the stack frame to be, it would overwrite other data on the stack. I realized this possibility while debugging a problem with calling complex Go code in a Go -> C -> Go cgo callback. This wasn't the bug I was looking for, it turns out, but I believe it is a real bug nonetheless. Switching to runtime.mcall, which only adds new frames to the stack and never jumps into functions running in existing ones, fixes this bug. * Move cgo-related code out of proc.c into cgocall.c. * Add very large comment describing cgo call sequences. * Simpilify, regularize cgo function implementations and names. * Add test suite as misc/cgo/test. Now the Go -> C path calls cgocall, which calls asmcgocall, and the C -> Go path calls cgocallback, which calls cgocallbackg. The shuffling, which affects mainly the callback case, moves most of the callback implementation to cgocallback running on the m->curg stack (not the m->g0 scheduler stack) and only while accounted for with $GOMAXPROCS (between calls to exitsyscall and entersyscall). The previous callback code did not block in startcgocallback's approximation to exitsyscall, so if, say, the garbage collector were running, it would still barge in and start doing things like call malloc. Similarly endcgocallback's approximation of entersyscall did not call matchmg to kick off new OS threads when necessary, which caused the bug in issue 1560. Fixes #1560. R=iant CC=golang-dev https://golang.org/cl/4253054
2011-03-07 08:37:42 -07:00
MOVW (g_sched+gobuf_sp)(g), SP
B runtime·oldstack(SB)
// void jmpdefer(fn, sp);
// called from deferreturn.
// 1. grab stored LR for caller
// 2. sub 4 bytes to get back to BL deferreturn
// 3. B to fn
TEXT runtime·jmpdefer(SB), 7, $0
MOVW 0(SP), LR
MOVW $-4(LR), LR // BL deferreturn
MOVW fn+0(FP), R7
MOVW argp+4(FP), SP
MOVW $-4(SP), SP // SP is 4 below argp, due to saved LR
MOVW 0(R7), R1
B (R1)
// Save state of caller into g->sched. Smashes R11.
TEXT gosave<>(SB),7,$0
MOVW LR, (g_sched+gobuf_pc)(g)
MOVW R13, (g_sched+gobuf_sp)(g)
MOVW $0, R11
MOVW R11, (g_sched+gobuf_lr)(g)
MOVW R11, (g_sched+gobuf_ret)(g)
MOVW R11, (g_sched+gobuf_ctxt)(g)
RET
// asmcgocall(void(*fn)(void*), void *arg)
// Call fn(arg) on the scheduler stack,
// aligned appropriately for the gcc ABI.
// See cgocall.c for more details.
runtime: scheduler, cgo reorganization * Change use of m->g0 stack (aka scheduler stack). * Provide runtime.mcall(f) to invoke f() on m->g0 stack. * Replace scheduler loop entry with runtime.mcall(schedule). Runtime.mcall eliminates the need for fake scheduler states that exist just to run a bit of code on the m->g0 stack (Grecovery, Gstackalloc). The elimination of the scheduler as a loop that stops and starts using gosave and gogo fixes a bad interaction with the way cgo uses the m->g0 stack. Cgo runs external (gcc-compiled) C functions on that stack, and then when calling back into Go, it sets m->g0->sched.sp below the added call frames, so that other uses of m->g0's stack will not interfere with those frames. Unfortunately, gogo (longjmp) back to the scheduler loop at this point would end up running scheduler with the lower sp, which no longer points at a valid stack frame for a call to scheduler. If scheduler then wrote any function call arguments or local variables to where it expected the stack frame to be, it would overwrite other data on the stack. I realized this possibility while debugging a problem with calling complex Go code in a Go -> C -> Go cgo callback. This wasn't the bug I was looking for, it turns out, but I believe it is a real bug nonetheless. Switching to runtime.mcall, which only adds new frames to the stack and never jumps into functions running in existing ones, fixes this bug. * Move cgo-related code out of proc.c into cgocall.c. * Add very large comment describing cgo call sequences. * Simpilify, regularize cgo function implementations and names. * Add test suite as misc/cgo/test. Now the Go -> C path calls cgocall, which calls asmcgocall, and the C -> Go path calls cgocallback, which calls cgocallbackg. The shuffling, which affects mainly the callback case, moves most of the callback implementation to cgocallback running on the m->curg stack (not the m->g0 scheduler stack) and only while accounted for with $GOMAXPROCS (between calls to exitsyscall and entersyscall). The previous callback code did not block in startcgocallback's approximation to exitsyscall, so if, say, the garbage collector were running, it would still barge in and start doing things like call malloc. Similarly endcgocallback's approximation of entersyscall did not call matchmg to kick off new OS threads when necessary, which caused the bug in issue 1560. Fixes #1560. R=iant CC=golang-dev https://golang.org/cl/4253054
2011-03-07 08:37:42 -07:00
TEXT runtime·asmcgocall(SB),7,$0
MOVW fn+0(FP), R1
MOVW arg+4(FP), R0
MOVW R13, R2
MOVW g, R5
// Figure out if we need to switch to m->g0 stack.
// We get called to create new OS threads too, and those
// come in on the m->g0 stack already.
MOVW m_g0(m), R3
CMP R3, g
BEQ 4(PC)
BL gosave<>(SB)
MOVW R3, g
MOVW (g_sched+gobuf_sp)(g), R13
// Now on a scheduling stack (a pthread-created stack).
SUB $24, R13
BIC $0x7, R13 // alignment for gcc ABI
MOVW R5, 20(R13) // save old g
MOVW R2, 16(R13) // save old SP
// R0 already contains the first argument
BL (R1)
// Restore registers, g, stack pointer.
MOVW 20(R13), g
MOVW 16(R13), R13
RET
// cgocallback(void (*fn)(void*), void *frame, uintptr framesize)
// Turn the fn into a Go func (by taking its address) and call
// cgocallback_gofunc.
TEXT runtime·cgocallback(SB),7,$12
MOVW $fn+0(FP), R0
MOVW R0, 4(R13)
MOVW frame+4(FP), R0
MOVW R0, 8(R13)
MOVW framesize+8(FP), R0
MOVW R0, 12(R13)
MOVW $runtime·cgocallback_gofunc(SB), R0
BL (R0)
RET
// cgocallback_gofunc(void (*fn)(void*), void *frame, uintptr framesize)
// See cgocall.c for more details.
TEXT runtime·cgocallback_gofunc(SB),7,$12
// Load m and g from thread-local storage.
MOVW _cgo_load_gm(SB), R0
CMP $0, R0
BL.NE (R0)
// If m is nil, Go did not create the current thread.
// Call needm to obtain one for temporary use.
// In this case, we're running on the thread stack, so there's
// lots of space, but the linker doesn't know. Hide the call from
// the linker analysis by using an indirect call.
MOVW m, savedm-12(SP)
CMP $0, m
B.NE havem
MOVW $runtime·needm(SB), R0
BL (R0)
havem:
// Now there's a valid m, and we're running on its m->g0.
// Save current m->g0->sched.sp on stack and then set it to SP.
// Save current sp in m->g0->sched.sp in preparation for
// switch back to m->curg stack.
MOVW m_g0(m), R3
MOVW (g_sched+gobuf_sp)(R3), R4
MOVW.W R4, -4(R13)
MOVW R13, (g_sched+gobuf_sp)(R3)
// Switch to m->curg stack and call runtime.cgocallbackg
// with the three arguments. Because we are taking over
// the execution of m->curg but *not* resuming what had
// been running, we need to save that information (m->curg->sched)
// so that we can restore it when we're done.
// We can restore m->curg->sched.sp easily, because calling
// runtime.cgocallbackg leaves SP unchanged upon return.
// To save m->curg->sched.pc, we push it onto the stack.
// This has the added benefit that it looks to the traceback
// routine like cgocallbackg is going to return to that
// PC (because we defined cgocallbackg to have
// a frame size of 12, the same amount that we use below),
// so that the traceback will seamlessly trace back into
// the earlier calls.
MOVW fn+4(FP), R0
MOVW frame+8(FP), R1
MOVW framesize+12(FP), R2
MOVW m_curg(m), g
MOVW (g_sched+gobuf_sp)(g), R4 // prepare stack as R4
// Push gobuf.pc
// Frame size here must match the frame size above plus the push
// to trick traceback routines into doing the right thing.
MOVW (g_sched+gobuf_pc)(g), R5
MOVW.W R5, -20(R4)
// Push arguments to cgocallbackg.
MOVW R0, 4(R4)
MOVW R1, 8(R4)
MOVW R2, 12(R4)
// Switch stack and make the call.
MOVW R4, R13
BL runtime·cgocallbackg(SB)
// Restore g->sched (== m->curg->sched) from saved values.
MOVW 0(R13), R5
MOVW R5, (g_sched+gobuf_pc)(g)
ADD $(16+4), R13, R4
MOVW R4, (g_sched+gobuf_sp)(g)
// Switch back to m->g0's stack and restore m->g0->sched.sp.
// (Unlike m->curg, the g0 goroutine never uses sched.pc,
// so we do not have to restore it.)
MOVW m_g0(m), g
MOVW (g_sched+gobuf_sp)(g), R13
// POP R6
MOVW 0(R13), R6
ADD $4, R13
MOVW R6, (g_sched+gobuf_sp)(g)
runtime: scheduler, cgo reorganization * Change use of m->g0 stack (aka scheduler stack). * Provide runtime.mcall(f) to invoke f() on m->g0 stack. * Replace scheduler loop entry with runtime.mcall(schedule). Runtime.mcall eliminates the need for fake scheduler states that exist just to run a bit of code on the m->g0 stack (Grecovery, Gstackalloc). The elimination of the scheduler as a loop that stops and starts using gosave and gogo fixes a bad interaction with the way cgo uses the m->g0 stack. Cgo runs external (gcc-compiled) C functions on that stack, and then when calling back into Go, it sets m->g0->sched.sp below the added call frames, so that other uses of m->g0's stack will not interfere with those frames. Unfortunately, gogo (longjmp) back to the scheduler loop at this point would end up running scheduler with the lower sp, which no longer points at a valid stack frame for a call to scheduler. If scheduler then wrote any function call arguments or local variables to where it expected the stack frame to be, it would overwrite other data on the stack. I realized this possibility while debugging a problem with calling complex Go code in a Go -> C -> Go cgo callback. This wasn't the bug I was looking for, it turns out, but I believe it is a real bug nonetheless. Switching to runtime.mcall, which only adds new frames to the stack and never jumps into functions running in existing ones, fixes this bug. * Move cgo-related code out of proc.c into cgocall.c. * Add very large comment describing cgo call sequences. * Simpilify, regularize cgo function implementations and names. * Add test suite as misc/cgo/test. Now the Go -> C path calls cgocall, which calls asmcgocall, and the C -> Go path calls cgocallback, which calls cgocallbackg. The shuffling, which affects mainly the callback case, moves most of the callback implementation to cgocallback running on the m->curg stack (not the m->g0 scheduler stack) and only while accounted for with $GOMAXPROCS (between calls to exitsyscall and entersyscall). The previous callback code did not block in startcgocallback's approximation to exitsyscall, so if, say, the garbage collector were running, it would still barge in and start doing things like call malloc. Similarly endcgocallback's approximation of entersyscall did not call matchmg to kick off new OS threads when necessary, which caused the bug in issue 1560. Fixes #1560. R=iant CC=golang-dev https://golang.org/cl/4253054
2011-03-07 08:37:42 -07:00
// If the m on entry was nil, we called needm above to borrow an m
// for the duration of the call. Since the call is over, return it with dropm.
MOVW savedm-12(SP), R6
CMP $0, R6
B.NE 3(PC)
MOVW $runtime·dropm(SB), R0
BL (R0)
// Done!
RET
runtime: scheduler, cgo reorganization * Change use of m->g0 stack (aka scheduler stack). * Provide runtime.mcall(f) to invoke f() on m->g0 stack. * Replace scheduler loop entry with runtime.mcall(schedule). Runtime.mcall eliminates the need for fake scheduler states that exist just to run a bit of code on the m->g0 stack (Grecovery, Gstackalloc). The elimination of the scheduler as a loop that stops and starts using gosave and gogo fixes a bad interaction with the way cgo uses the m->g0 stack. Cgo runs external (gcc-compiled) C functions on that stack, and then when calling back into Go, it sets m->g0->sched.sp below the added call frames, so that other uses of m->g0's stack will not interfere with those frames. Unfortunately, gogo (longjmp) back to the scheduler loop at this point would end up running scheduler with the lower sp, which no longer points at a valid stack frame for a call to scheduler. If scheduler then wrote any function call arguments or local variables to where it expected the stack frame to be, it would overwrite other data on the stack. I realized this possibility while debugging a problem with calling complex Go code in a Go -> C -> Go cgo callback. This wasn't the bug I was looking for, it turns out, but I believe it is a real bug nonetheless. Switching to runtime.mcall, which only adds new frames to the stack and never jumps into functions running in existing ones, fixes this bug. * Move cgo-related code out of proc.c into cgocall.c. * Add very large comment describing cgo call sequences. * Simpilify, regularize cgo function implementations and names. * Add test suite as misc/cgo/test. Now the Go -> C path calls cgocall, which calls asmcgocall, and the C -> Go path calls cgocallback, which calls cgocallbackg. The shuffling, which affects mainly the callback case, moves most of the callback implementation to cgocallback running on the m->curg stack (not the m->g0 scheduler stack) and only while accounted for with $GOMAXPROCS (between calls to exitsyscall and entersyscall). The previous callback code did not block in startcgocallback's approximation to exitsyscall, so if, say, the garbage collector were running, it would still barge in and start doing things like call malloc. Similarly endcgocallback's approximation of entersyscall did not call matchmg to kick off new OS threads when necessary, which caused the bug in issue 1560. Fixes #1560. R=iant CC=golang-dev https://golang.org/cl/4253054
2011-03-07 08:37:42 -07:00
// void setmg(M*, G*); set m and g. for use by needm.
TEXT runtime·setmg(SB), 7, $0
MOVW mm+0(FP), m
MOVW gg+4(FP), g
// Save m and g to thread-local storage.
MOVW _cgo_save_gm(SB), R0
CMP $0, R0
BL.NE (R0)
RET
TEXT runtime·getcallerpc(SB),7,$-4
MOVW 0(SP), R0
RET
TEXT runtime·setcallerpc(SB),7,$-4
MOVW x+4(FP), R0
MOVW R0, 0(SP)
RET
TEXT runtime·getcallersp(SB),7,$-4
MOVW 0(FP), R0
MOVW $-4(R0), R0
RET
TEXT runtime·emptyfunc(SB),0,$0
RET
TEXT runtime·abort(SB),7,$-4
MOVW $0, R0
MOVW (R0), R1
// bool armcas(int32 *val, int32 old, int32 new)
// Atomically:
// if(*val == old){
// *val = new;
// return 1;
// }else
// return 0;
//
// To implement runtime·cas in sys_$GOOS_arm.s
// using the native instructions, use:
//
// TEXT runtime·cas(SB),7,$0
// B runtime·armcas(SB)
//
TEXT runtime·armcas(SB),7,$0
MOVW valptr+0(FP), R1
MOVW old+4(FP), R2
MOVW new+8(FP), R3
casl:
LDREX (R1), R0
CMP R0, R2
BNE casfail
STREX R3, (R1), R0
CMP $0, R0
BNE casl
MOVW $1, R0
RET
casfail:
MOVW $0, R0
RET
TEXT runtime·stackguard(SB),7,$0
MOVW R13, R1
MOVW g_stackguard(g), R2
MOVW R1, sp+0(FP)
MOVW R2, limit+4(FP)
RET
// AES hashing not implemented for ARM
TEXT runtime·aeshash(SB),7,$-4
MOVW $0, R0
MOVW (R0), R1
TEXT runtime·aeshash32(SB),7,$-4
MOVW $0, R0
MOVW (R0), R1
TEXT runtime·aeshash64(SB),7,$-4
MOVW $0, R0
MOVW (R0), R1
TEXT runtime·aeshashstr(SB),7,$-4
MOVW $0, R0
MOVW (R0), R1
TEXT runtime·memeq(SB),7,$-4
MOVW a+0(FP), R1
MOVW b+4(FP), R2
MOVW n+8(FP), R3
ADD R1, R3, R6
MOVW $1, R0
_next:
CMP R1, R6
RET.EQ
MOVBU.P 1(R1), R4
MOVBU.P 1(R2), R5
CMP R4, R5
BEQ _next
MOVW $0, R0
RET