delve/pkg/proc/moduledata.go

206 lines
4.7 KiB
Go
Raw Normal View History

package proc
import (
"go/constant"
"unsafe"
)
// delve counterpart to runtime.moduledata
type moduleData struct {
text, etext uint64
types, etypes uint64
typemapVar *Variable
}
proc: support debugging plugins (#1414) This change splits the BinaryInfo object into a slice of Image objects containing information about the base executable and each loaded shared library (note: go plugins are shared libraries). Delve backens are supposed to call BinaryInfo.AddImage whenever they detect that a new shared library has been loaded. Member fields of BinaryInfo that are used to speed up access to dwarf (Functions, packageVars, consts, etc...) remain part of BinaryInfo and are updated to reference the correct image object. This simplifies this change. This approach has a few shortcomings: 1. Multiple shared libraries can define functions or globals with the same name and we have no way to disambiguate between them. 2. We don't have a way to handle library unloading. Both of those affect C shared libraries much more than they affect go plugins. Go plugins can't be unloaded at all and a lot of name collisions are prevented by import paths. There's only one problem that is concerning: if two plugins both import the same package they will end up with multiple definition for the same function. For example if two plugins use fmt.Printf the final in-memory image (and therefore our BinaryInfo object) will end up with two copies of fmt.Printf at different memory addresses. If a user types break fmt.Printf a breakpoint should be created at *both* locations. Allowing this is a relatively complex change that should be done in a different PR than this. For this reason I consider this approach an acceptable and sustainable stopgap. Updates #865
2019-05-08 21:06:38 +00:00
func loadModuleData(bi *BinaryInfo, mem MemoryReadWriter) ([]moduleData, error) {
scope := globalScope(bi, bi.Images[0], mem)
var md *Variable
md, err := scope.findGlobal("runtime", "firstmoduledata")
proc: support debugging plugins (#1414) This change splits the BinaryInfo object into a slice of Image objects containing information about the base executable and each loaded shared library (note: go plugins are shared libraries). Delve backens are supposed to call BinaryInfo.AddImage whenever they detect that a new shared library has been loaded. Member fields of BinaryInfo that are used to speed up access to dwarf (Functions, packageVars, consts, etc...) remain part of BinaryInfo and are updated to reference the correct image object. This simplifies this change. This approach has a few shortcomings: 1. Multiple shared libraries can define functions or globals with the same name and we have no way to disambiguate between them. 2. We don't have a way to handle library unloading. Both of those affect C shared libraries much more than they affect go plugins. Go plugins can't be unloaded at all and a lot of name collisions are prevented by import paths. There's only one problem that is concerning: if two plugins both import the same package they will end up with multiple definition for the same function. For example if two plugins use fmt.Printf the final in-memory image (and therefore our BinaryInfo object) will end up with two copies of fmt.Printf at different memory addresses. If a user types break fmt.Printf a breakpoint should be created at *both* locations. Allowing this is a relatively complex change that should be done in a different PR than this. For this reason I consider this approach an acceptable and sustainable stopgap. Updates #865
2019-05-08 21:06:38 +00:00
if err != nil {
return nil, err
}
proc: support debugging plugins (#1414) This change splits the BinaryInfo object into a slice of Image objects containing information about the base executable and each loaded shared library (note: go plugins are shared libraries). Delve backens are supposed to call BinaryInfo.AddImage whenever they detect that a new shared library has been loaded. Member fields of BinaryInfo that are used to speed up access to dwarf (Functions, packageVars, consts, etc...) remain part of BinaryInfo and are updated to reference the correct image object. This simplifies this change. This approach has a few shortcomings: 1. Multiple shared libraries can define functions or globals with the same name and we have no way to disambiguate between them. 2. We don't have a way to handle library unloading. Both of those affect C shared libraries much more than they affect go plugins. Go plugins can't be unloaded at all and a lot of name collisions are prevented by import paths. There's only one problem that is concerning: if two plugins both import the same package they will end up with multiple definition for the same function. For example if two plugins use fmt.Printf the final in-memory image (and therefore our BinaryInfo object) will end up with two copies of fmt.Printf at different memory addresses. If a user types break fmt.Printf a breakpoint should be created at *both* locations. Allowing this is a relatively complex change that should be done in a different PR than this. For this reason I consider this approach an acceptable and sustainable stopgap. Updates #865
2019-05-08 21:06:38 +00:00
r := []moduleData{}
for md.Addr != 0 {
const (
typesField = "types"
etypesField = "etypes"
textField = "text"
etextField = "etext"
nextField = "next"
typemapField = "typemap"
)
vars := map[string]*Variable{}
for _, fieldName := range []string{typesField, etypesField, textField, etextField, nextField, typemapField} {
var err error
vars[fieldName], err = md.structMember(fieldName)
if err != nil {
return nil, err
}
proc: support debugging plugins (#1414) This change splits the BinaryInfo object into a slice of Image objects containing information about the base executable and each loaded shared library (note: go plugins are shared libraries). Delve backens are supposed to call BinaryInfo.AddImage whenever they detect that a new shared library has been loaded. Member fields of BinaryInfo that are used to speed up access to dwarf (Functions, packageVars, consts, etc...) remain part of BinaryInfo and are updated to reference the correct image object. This simplifies this change. This approach has a few shortcomings: 1. Multiple shared libraries can define functions or globals with the same name and we have no way to disambiguate between them. 2. We don't have a way to handle library unloading. Both of those affect C shared libraries much more than they affect go plugins. Go plugins can't be unloaded at all and a lot of name collisions are prevented by import paths. There's only one problem that is concerning: if two plugins both import the same package they will end up with multiple definition for the same function. For example if two plugins use fmt.Printf the final in-memory image (and therefore our BinaryInfo object) will end up with two copies of fmt.Printf at different memory addresses. If a user types break fmt.Printf a breakpoint should be created at *both* locations. Allowing this is a relatively complex change that should be done in a different PR than this. For this reason I consider this approach an acceptable and sustainable stopgap. Updates #865
2019-05-08 21:06:38 +00:00
}
var err error
touint := func(name string) (ret uint64) {
proc: support debugging plugins (#1414) This change splits the BinaryInfo object into a slice of Image objects containing information about the base executable and each loaded shared library (note: go plugins are shared libraries). Delve backens are supposed to call BinaryInfo.AddImage whenever they detect that a new shared library has been loaded. Member fields of BinaryInfo that are used to speed up access to dwarf (Functions, packageVars, consts, etc...) remain part of BinaryInfo and are updated to reference the correct image object. This simplifies this change. This approach has a few shortcomings: 1. Multiple shared libraries can define functions or globals with the same name and we have no way to disambiguate between them. 2. We don't have a way to handle library unloading. Both of those affect C shared libraries much more than they affect go plugins. Go plugins can't be unloaded at all and a lot of name collisions are prevented by import paths. There's only one problem that is concerning: if two plugins both import the same package they will end up with multiple definition for the same function. For example if two plugins use fmt.Printf the final in-memory image (and therefore our BinaryInfo object) will end up with two copies of fmt.Printf at different memory addresses. If a user types break fmt.Printf a breakpoint should be created at *both* locations. Allowing this is a relatively complex change that should be done in a different PR than this. For this reason I consider this approach an acceptable and sustainable stopgap. Updates #865
2019-05-08 21:06:38 +00:00
if err == nil {
var n uint64
n, err = vars[name].asUint()
ret = n
}
proc: support debugging plugins (#1414) This change splits the BinaryInfo object into a slice of Image objects containing information about the base executable and each loaded shared library (note: go plugins are shared libraries). Delve backens are supposed to call BinaryInfo.AddImage whenever they detect that a new shared library has been loaded. Member fields of BinaryInfo that are used to speed up access to dwarf (Functions, packageVars, consts, etc...) remain part of BinaryInfo and are updated to reference the correct image object. This simplifies this change. This approach has a few shortcomings: 1. Multiple shared libraries can define functions or globals with the same name and we have no way to disambiguate between them. 2. We don't have a way to handle library unloading. Both of those affect C shared libraries much more than they affect go plugins. Go plugins can't be unloaded at all and a lot of name collisions are prevented by import paths. There's only one problem that is concerning: if two plugins both import the same package they will end up with multiple definition for the same function. For example if two plugins use fmt.Printf the final in-memory image (and therefore our BinaryInfo object) will end up with two copies of fmt.Printf at different memory addresses. If a user types break fmt.Printf a breakpoint should be created at *both* locations. Allowing this is a relatively complex change that should be done in a different PR than this. For this reason I consider this approach an acceptable and sustainable stopgap. Updates #865
2019-05-08 21:06:38 +00:00
return ret
}
proc: support debugging plugins (#1414) This change splits the BinaryInfo object into a slice of Image objects containing information about the base executable and each loaded shared library (note: go plugins are shared libraries). Delve backens are supposed to call BinaryInfo.AddImage whenever they detect that a new shared library has been loaded. Member fields of BinaryInfo that are used to speed up access to dwarf (Functions, packageVars, consts, etc...) remain part of BinaryInfo and are updated to reference the correct image object. This simplifies this change. This approach has a few shortcomings: 1. Multiple shared libraries can define functions or globals with the same name and we have no way to disambiguate between them. 2. We don't have a way to handle library unloading. Both of those affect C shared libraries much more than they affect go plugins. Go plugins can't be unloaded at all and a lot of name collisions are prevented by import paths. There's only one problem that is concerning: if two plugins both import the same package they will end up with multiple definition for the same function. For example if two plugins use fmt.Printf the final in-memory image (and therefore our BinaryInfo object) will end up with two copies of fmt.Printf at different memory addresses. If a user types break fmt.Printf a breakpoint should be created at *both* locations. Allowing this is a relatively complex change that should be done in a different PR than this. For this reason I consider this approach an acceptable and sustainable stopgap. Updates #865
2019-05-08 21:06:38 +00:00
r = append(r, moduleData{
types: touint(typesField), etypes: touint(etypesField),
text: touint(textField), etext: touint(etextField),
typemapVar: vars[typemapField],
})
if err != nil {
return nil, err
}
proc: support debugging plugins (#1414) This change splits the BinaryInfo object into a slice of Image objects containing information about the base executable and each loaded shared library (note: go plugins are shared libraries). Delve backens are supposed to call BinaryInfo.AddImage whenever they detect that a new shared library has been loaded. Member fields of BinaryInfo that are used to speed up access to dwarf (Functions, packageVars, consts, etc...) remain part of BinaryInfo and are updated to reference the correct image object. This simplifies this change. This approach has a few shortcomings: 1. Multiple shared libraries can define functions or globals with the same name and we have no way to disambiguate between them. 2. We don't have a way to handle library unloading. Both of those affect C shared libraries much more than they affect go plugins. Go plugins can't be unloaded at all and a lot of name collisions are prevented by import paths. There's only one problem that is concerning: if two plugins both import the same package they will end up with multiple definition for the same function. For example if two plugins use fmt.Printf the final in-memory image (and therefore our BinaryInfo object) will end up with two copies of fmt.Printf at different memory addresses. If a user types break fmt.Printf a breakpoint should be created at *both* locations. Allowing this is a relatively complex change that should be done in a different PR than this. For this reason I consider this approach an acceptable and sustainable stopgap. Updates #865
2019-05-08 21:06:38 +00:00
md = vars[nextField].maybeDereference()
if md.Unreadable != nil {
return nil, md.Unreadable
}
}
proc: support debugging plugins (#1414) This change splits the BinaryInfo object into a slice of Image objects containing information about the base executable and each loaded shared library (note: go plugins are shared libraries). Delve backens are supposed to call BinaryInfo.AddImage whenever they detect that a new shared library has been loaded. Member fields of BinaryInfo that are used to speed up access to dwarf (Functions, packageVars, consts, etc...) remain part of BinaryInfo and are updated to reference the correct image object. This simplifies this change. This approach has a few shortcomings: 1. Multiple shared libraries can define functions or globals with the same name and we have no way to disambiguate between them. 2. We don't have a way to handle library unloading. Both of those affect C shared libraries much more than they affect go plugins. Go plugins can't be unloaded at all and a lot of name collisions are prevented by import paths. There's only one problem that is concerning: if two plugins both import the same package they will end up with multiple definition for the same function. For example if two plugins use fmt.Printf the final in-memory image (and therefore our BinaryInfo object) will end up with two copies of fmt.Printf at different memory addresses. If a user types break fmt.Printf a breakpoint should be created at *both* locations. Allowing this is a relatively complex change that should be done in a different PR than this. For this reason I consider this approach an acceptable and sustainable stopgap. Updates #865
2019-05-08 21:06:38 +00:00
return r, nil
}
func findModuleDataForType(bi *BinaryInfo, mds []moduleData, typeAddr uint64, mem MemoryReadWriter) *moduleData {
proc: support debugging plugins (#1414) This change splits the BinaryInfo object into a slice of Image objects containing information about the base executable and each loaded shared library (note: go plugins are shared libraries). Delve backens are supposed to call BinaryInfo.AddImage whenever they detect that a new shared library has been loaded. Member fields of BinaryInfo that are used to speed up access to dwarf (Functions, packageVars, consts, etc...) remain part of BinaryInfo and are updated to reference the correct image object. This simplifies this change. This approach has a few shortcomings: 1. Multiple shared libraries can define functions or globals with the same name and we have no way to disambiguate between them. 2. We don't have a way to handle library unloading. Both of those affect C shared libraries much more than they affect go plugins. Go plugins can't be unloaded at all and a lot of name collisions are prevented by import paths. There's only one problem that is concerning: if two plugins both import the same package they will end up with multiple definition for the same function. For example if two plugins use fmt.Printf the final in-memory image (and therefore our BinaryInfo object) will end up with two copies of fmt.Printf at different memory addresses. If a user types break fmt.Printf a breakpoint should be created at *both* locations. Allowing this is a relatively complex change that should be done in a different PR than this. For this reason I consider this approach an acceptable and sustainable stopgap. Updates #865
2019-05-08 21:06:38 +00:00
for i := range mds {
if typeAddr >= mds[i].types && typeAddr < mds[i].etypes {
return &mds[i]
}
}
proc: support debugging plugins (#1414) This change splits the BinaryInfo object into a slice of Image objects containing information about the base executable and each loaded shared library (note: go plugins are shared libraries). Delve backens are supposed to call BinaryInfo.AddImage whenever they detect that a new shared library has been loaded. Member fields of BinaryInfo that are used to speed up access to dwarf (Functions, packageVars, consts, etc...) remain part of BinaryInfo and are updated to reference the correct image object. This simplifies this change. This approach has a few shortcomings: 1. Multiple shared libraries can define functions or globals with the same name and we have no way to disambiguate between them. 2. We don't have a way to handle library unloading. Both of those affect C shared libraries much more than they affect go plugins. Go plugins can't be unloaded at all and a lot of name collisions are prevented by import paths. There's only one problem that is concerning: if two plugins both import the same package they will end up with multiple definition for the same function. For example if two plugins use fmt.Printf the final in-memory image (and therefore our BinaryInfo object) will end up with two copies of fmt.Printf at different memory addresses. If a user types break fmt.Printf a breakpoint should be created at *both* locations. Allowing this is a relatively complex change that should be done in a different PR than this. For this reason I consider this approach an acceptable and sustainable stopgap. Updates #865
2019-05-08 21:06:38 +00:00
return nil
}
func resolveTypeOff(bi *BinaryInfo, mds []moduleData, typeAddr, off uint64, mem MemoryReadWriter) (*Variable, error) {
// See runtime.(*_type).typeOff in $GOROOT/src/runtime/type.go
proc: support debugging plugins (#1414) This change splits the BinaryInfo object into a slice of Image objects containing information about the base executable and each loaded shared library (note: go plugins are shared libraries). Delve backens are supposed to call BinaryInfo.AddImage whenever they detect that a new shared library has been loaded. Member fields of BinaryInfo that are used to speed up access to dwarf (Functions, packageVars, consts, etc...) remain part of BinaryInfo and are updated to reference the correct image object. This simplifies this change. This approach has a few shortcomings: 1. Multiple shared libraries can define functions or globals with the same name and we have no way to disambiguate between them. 2. We don't have a way to handle library unloading. Both of those affect C shared libraries much more than they affect go plugins. Go plugins can't be unloaded at all and a lot of name collisions are prevented by import paths. There's only one problem that is concerning: if two plugins both import the same package they will end up with multiple definition for the same function. For example if two plugins use fmt.Printf the final in-memory image (and therefore our BinaryInfo object) will end up with two copies of fmt.Printf at different memory addresses. If a user types break fmt.Printf a breakpoint should be created at *both* locations. Allowing this is a relatively complex change that should be done in a different PR than this. For this reason I consider this approach an acceptable and sustainable stopgap. Updates #865
2019-05-08 21:06:38 +00:00
md := findModuleDataForType(bi, mds, typeAddr, mem)
rtyp, err := bi.findType("runtime._type")
if err != nil {
return nil, err
}
if md == nil {
v, err := reflectOffsMapAccess(bi, off, mem)
if err != nil {
return nil, err
}
proc: Improve performance of loadMap on very large sparse maps Users can create sparse maps in two ways, either by: a) adding lots of entries to a map and then deleting most of them, or b) using the make(mapType, N) expression with a very large N When this happens reading the resulting map will be very slow because loadMap needs to scan many buckets for each entry it finds. Technically this is not a bug, the user just created a map that's very sparse and therefore very slow to read. However it's very annoying to have the debugger hang for several seconds when trying to read the local variables just because one of them (which you might not even be interested into) happens to be a very sparse map. There is an easy mitigation to this problem: not reading any additional buckets once we know that we have already read all entries of the map, or as many entries as we need to fulfill the MaxArrayValues parameter. Unfortunately this is mostly useless, a VLSM (Very Large Sparse Map) with a single entry will still be slow to access, because the single entry in the map could easily end up in the last bucket. The obvious solution to this problem is to set a limit to the number of buckets we read when loading a map. However there is no good way to set this limit. If we hardcode it there will be no way to print maps that are beyond whatever limit we pick. We could let users (or clients) specify it but the meaning of such knob would be arcane and they would have no way of picking a good value (because there is no objectively good value for it). The solution used in this commit is to set an arbirtray limit on the number of buckets we read but only when loadMap is invoked through API calls ListLocalVars and ListFunctionArgs. In this way `ListLocalVars` and `ListFunctionArgs` (which are often invoked automatically by GUI clients) remain fast even in presence of a VLSM, but the contents of the VLSM can still be inspected using `EvalVariable`.
2018-10-29 11:22:03 +00:00
v.loadValue(LoadConfig{false, 1, 0, 0, -1, 0})
addr, _ := constant.Int64Val(v.Value)
return v.newVariable(v.Name, uint64(addr), rtyp, mem), nil
}
if t, _ := md.typemapVar.mapAccess(newConstant(constant.MakeUint64(uint64(off)), mem)); t != nil {
return t, nil
}
res := md.types + off
return newVariable("", uint64(res), rtyp, bi, mem), nil
}
func resolveNameOff(bi *BinaryInfo, mds []moduleData, typeAddr, off uint64, mem MemoryReadWriter) (name, tag string, pkgpathoff int32, err error) {
// See runtime.resolveNameOff in $GOROOT/src/runtime/type.go
proc: support debugging plugins (#1414) This change splits the BinaryInfo object into a slice of Image objects containing information about the base executable and each loaded shared library (note: go plugins are shared libraries). Delve backens are supposed to call BinaryInfo.AddImage whenever they detect that a new shared library has been loaded. Member fields of BinaryInfo that are used to speed up access to dwarf (Functions, packageVars, consts, etc...) remain part of BinaryInfo and are updated to reference the correct image object. This simplifies this change. This approach has a few shortcomings: 1. Multiple shared libraries can define functions or globals with the same name and we have no way to disambiguate between them. 2. We don't have a way to handle library unloading. Both of those affect C shared libraries much more than they affect go plugins. Go plugins can't be unloaded at all and a lot of name collisions are prevented by import paths. There's only one problem that is concerning: if two plugins both import the same package they will end up with multiple definition for the same function. For example if two plugins use fmt.Printf the final in-memory image (and therefore our BinaryInfo object) will end up with two copies of fmt.Printf at different memory addresses. If a user types break fmt.Printf a breakpoint should be created at *both* locations. Allowing this is a relatively complex change that should be done in a different PR than this. For this reason I consider this approach an acceptable and sustainable stopgap. Updates #865
2019-05-08 21:06:38 +00:00
for _, md := range mds {
if typeAddr >= md.types && typeAddr < md.etypes {
return loadName(bi, md.types+off, mem)
}
}
v, err := reflectOffsMapAccess(bi, off, mem)
if err != nil {
return "", "", 0, err
}
resv := v.maybeDereference()
if resv.Unreadable != nil {
return "", "", 0, resv.Unreadable
}
return loadName(bi, resv.Addr, mem)
}
func reflectOffsMapAccess(bi *BinaryInfo, off uint64, mem MemoryReadWriter) (*Variable, error) {
proc: support debugging plugins (#1414) This change splits the BinaryInfo object into a slice of Image objects containing information about the base executable and each loaded shared library (note: go plugins are shared libraries). Delve backens are supposed to call BinaryInfo.AddImage whenever they detect that a new shared library has been loaded. Member fields of BinaryInfo that are used to speed up access to dwarf (Functions, packageVars, consts, etc...) remain part of BinaryInfo and are updated to reference the correct image object. This simplifies this change. This approach has a few shortcomings: 1. Multiple shared libraries can define functions or globals with the same name and we have no way to disambiguate between them. 2. We don't have a way to handle library unloading. Both of those affect C shared libraries much more than they affect go plugins. Go plugins can't be unloaded at all and a lot of name collisions are prevented by import paths. There's only one problem that is concerning: if two plugins both import the same package they will end up with multiple definition for the same function. For example if two plugins use fmt.Printf the final in-memory image (and therefore our BinaryInfo object) will end up with two copies of fmt.Printf at different memory addresses. If a user types break fmt.Printf a breakpoint should be created at *both* locations. Allowing this is a relatively complex change that should be done in a different PR than this. For this reason I consider this approach an acceptable and sustainable stopgap. Updates #865
2019-05-08 21:06:38 +00:00
scope := globalScope(bi, bi.Images[0], mem)
reflectOffs, err := scope.findGlobal("runtime", "reflectOffs")
if err != nil {
return nil, err
}
reflectOffsm, err := reflectOffs.structMember("m")
if err != nil {
return nil, err
}
return reflectOffsm.mapAccess(newConstant(constant.MakeUint64(uint64(off)), mem))
}
const (
// flags for the name struct (see 'type name struct' in $GOROOT/src/reflect/type.go)
nameflagExported = 1 << 0
nameflagHasTag = 1 << 1
nameflagHasPkg = 1 << 2
)
func loadName(bi *BinaryInfo, addr uint64, mem MemoryReadWriter) (name, tag string, pkgpathoff int32, err error) {
off := addr
namedata := make([]byte, 3)
_, err = mem.ReadMemory(namedata, off)
off += 3
if err != nil {
return "", "", 0, err
}
namelen := uint16(namedata[1])<<8 | uint16(namedata[2])
rawstr := make([]byte, int(namelen))
_, err = mem.ReadMemory(rawstr, off)
off += uint64(namelen)
if err != nil {
return "", "", 0, err
}
name = string(rawstr)
if namedata[0]&nameflagHasTag != 0 {
taglendata := make([]byte, 2)
_, err = mem.ReadMemory(taglendata, off)
off += 2
if err != nil {
return "", "", 0, err
}
taglen := uint16(taglendata[0])<<8 | uint16(taglendata[1])
rawstr := make([]byte, int(taglen))
_, err = mem.ReadMemory(rawstr, off)
off += uint64(taglen)
if err != nil {
return "", "", 0, err
}
tag = string(rawstr)
}
if namedata[0]&nameflagHasPkg != 0 {
pkgdata := make([]byte, 4)
_, err = mem.ReadMemory(pkgdata, off)
if err != nil {
return "", "", 0, err
}
// see func pkgPath in $GOROOT/src/reflect/type.go
copy((*[4]byte)(unsafe.Pointer(&pkgpathoff))[:], pkgdata)
}
return name, tag, pkgpathoff, nil
}