llvm-project/lldb/source/Host/macosx/Host.mm

1504 lines
49 KiB
Plaintext
Raw Normal View History

//===-- Host.mm -------------------------------------------------*- C++ -*-===//
//
// The LLVM Compiler Infrastructure
//
// This file is distributed under the University of Illinois Open Source
// License. See LICENSE.TXT for details.
//
//===----------------------------------------------------------------------===//
#include "lldb/Host/Host.h"
Added a new Host call to find LLDB related paths: static bool Host::GetLLDBPath (lldb::PathType path_type, FileSpec &file_spec); This will fill in "file_spec" with an appropriate path that is appropriate for the current Host OS. MacOSX will return paths within the LLDB.framework, and other unixes will return the paths they want. The current PathType enums are: typedef enum PathType { ePathTypeLLDBShlibDir, // The directory where the lldb.so (unix) or LLDB mach-o file in LLDB.framework (MacOSX) exists ePathTypeSupportExecutableDir, // Find LLDB support executable directory (debugserver, etc) ePathTypeHeaderDir, // Find LLDB header file directory ePathTypePythonDir // Find Python modules (PYTHONPATH) directory } PathType; All places that were finding executables are and python paths are now updated to use this Host call. Added another new host call to launch the inferior in a terminal. This ability will be very host specific and doesn't need to be supported on all systems. MacOSX currently will create a new .command file and tell Terminal.app to open the .command file. It also uses the new "darwin-debug" app which is a small app that uses posix to exec (no fork) and stop at the entry point of the program. The GDB remote plug-in is almost able launch a process and attach to it, it currently will spawn the process, but it won't attach to it just yet. This will let LLDB not have to share the terminal with another process and a new terminal window will pop up when you launch. This won't get hooked up until we work out all of the kinks. The new Host function is: static lldb::pid_t Host::LaunchInNewTerminal ( const char **argv, // argv[0] is executable const char **envp, const ArchSpec *arch_spec, bool stop_at_entry, bool disable_aslr); Cleaned up FileSpec::GetPath to not use strncpy() as it was always zero filling the entire path buffer. Fixed an issue with the dynamic checker function where I missed a '$' prefix that should have been added. llvm-svn: 116690
2010-10-18 06:03:32 +08:00
#include <AvailabilityMacros.h>
#if !defined(MAC_OS_X_VERSION_10_7) || \
MAC_OS_X_VERSION_MAX_ALLOWED < MAC_OS_X_VERSION_10_7
#define NO_XPC_SERVICES 1
#endif
#if !defined(NO_XPC_SERVICES)
#define __XPC_PRIVATE_H__
#include <xpc/xpc.h>
#define LaunchUsingXPCRightName "com.apple.dt.Xcode.RootDebuggingXPCService"
// These XPC messaging keys are used for communication between Host.mm and the
// XPC service.
#define LauncherXPCServiceAuthKey "auth-key"
#define LauncherXPCServiceArgPrefxKey "arg"
#define LauncherXPCServiceEnvPrefxKey "env"
#define LauncherXPCServiceCPUTypeKey "cpuType"
#define LauncherXPCServicePosixspawnFlagsKey "posixspawnFlags"
#define LauncherXPCServiceStdInPathKeyKey "stdInPath"
#define LauncherXPCServiceStdOutPathKeyKey "stdOutPath"
#define LauncherXPCServiceStdErrPathKeyKey "stdErrPath"
#define LauncherXPCServiceChildPIDKey "childPID"
#define LauncherXPCServiceErrorTypeKey "errorType"
#define LauncherXPCServiceCodeTypeKey "errorCode"
#endif
#include "llvm/Support/Host.h"
#include <asl.h>
#include <crt_externs.h>
#include <grp.h>
#include <libproc.h>
#include <pwd.h>
Moved the execution context that was in the Debugger into the CommandInterpreter where it was always being used. Make sure that Modules can track their object file offsets correctly to allow opening of sub object files (like the "__commpage" on darwin). Modified the Platforms to be able to launch processes. The first part of this move is the platform soon will become the entity that launches your program and when it does, it uses a new ProcessLaunchInfo class which encapsulates all process launching settings. This simplifies the internal APIs needed for launching. I want to slowly phase out process launching from the process classes, so for now we can still launch just as we used to, but eventually the platform is the object that should do the launching. Modified the Host::LaunchProcess in the MacOSX Host.mm to correctly be able to launch processes with all of the new eLaunchFlag settings. Modified any code that was manually launching processes to use the Host::LaunchProcess functions. Fixed an issue where lldb_private::Args had implicitly defined copy constructors that could do the wrong thing. This has now been fixed by adding an appropriate copy constructor and assignment operator. Make sure we don't add empty ModuleSP entries to a module list. Fixed the commpage module creation on MacOSX, but we still need to train the MacOSX dynamic loader to not get rid of it when it doesn't have an entry in the all image infos. Abstracted many more calls from in ProcessGDBRemote down into the GDBRemoteCommunicationClient subclass to make the classes cleaner and more efficient. Fixed the default iOS ARM register context to be correct and also added support for targets that don't support the qThreadStopInfo packet by selecting the current thread (only if needed) and then sending a stop reply packet. Debugserver can now start up with a --unix-socket (-u for short) and can then bind to port zero and send the port it bound to to a listening process on the other end. This allows the GDB remote platform to spawn new GDB server instances (debugserver) to allow platform debugging. llvm-svn: 129351
2011-04-12 13:54:46 +08:00
#include <spawn.h>
Fixed a race condition that could cause ProcessGDBRemote::DoResume() to return an error saying the resume timed out. Previously the thread that was trying to resume the process would eventually call ProcessGDBRemote::DoResume() which would broadcast an event over to the async GDB remote thread which would sent the continue packet to the remote gdb server. Right after this was sent, it would set a predicate boolean value (protected by a mutex and condition) and then the thread that issued the ProcessGDBRemote::DoResume() would then wait for that condition variable to be set. If the async gdb thread was too quick though, the predicate boolean value could have been set to true and back to false by the time the thread that issued the ProcessGDBRemote::DoResume() checks the boolean value. So we can't use the predicate value as a handshake. I have changed the code over to using a Event by having the GDB remote communication object post an event: GDBRemoteCommunication::eBroadcastBitRunPacketSent This allows reliable handshaking between the two threads and avoids the erroneous ProcessGDBRemote::DoResume() errors. Added a host backtrace service to allow in process backtraces when trying to track down tricky issues. I need to see if LLVM has any backtracing abilities abstracted in it already, and if so, use that, but I needed something ASAP for the current issue I was working on. The static function is: void Host::Backtrace (Stream &strm, uint32_t max_frames); And it will backtrace at most "max_frames" frames for the current thread and can be used with any of the Stream subclasses for logging. llvm-svn: 120793
2010-12-03 14:02:24 +08:00
#include <stdio.h>
#include <stdlib.h>
#include <sys/proc.h>
Added a new Host call to find LLDB related paths: static bool Host::GetLLDBPath (lldb::PathType path_type, FileSpec &file_spec); This will fill in "file_spec" with an appropriate path that is appropriate for the current Host OS. MacOSX will return paths within the LLDB.framework, and other unixes will return the paths they want. The current PathType enums are: typedef enum PathType { ePathTypeLLDBShlibDir, // The directory where the lldb.so (unix) or LLDB mach-o file in LLDB.framework (MacOSX) exists ePathTypeSupportExecutableDir, // Find LLDB support executable directory (debugserver, etc) ePathTypeHeaderDir, // Find LLDB header file directory ePathTypePythonDir // Find Python modules (PYTHONPATH) directory } PathType; All places that were finding executables are and python paths are now updated to use this Host call. Added another new host call to launch the inferior in a terminal. This ability will be very host specific and doesn't need to be supported on all systems. MacOSX currently will create a new .command file and tell Terminal.app to open the .command file. It also uses the new "darwin-debug" app which is a small app that uses posix to exec (no fork) and stop at the entry point of the program. The GDB remote plug-in is almost able launch a process and attach to it, it currently will spawn the process, but it won't attach to it just yet. This will let LLDB not have to share the terminal with another process and a new terminal window will pop up when you launch. This won't get hooked up until we work out all of the kinks. The new Host function is: static lldb::pid_t Host::LaunchInNewTerminal ( const char **argv, // argv[0] is executable const char **envp, const ArchSpec *arch_spec, bool stop_at_entry, bool disable_aslr); Cleaned up FileSpec::GetPath to not use strncpy() as it was always zero filling the entire path buffer. Fixed an issue with the dynamic checker function where I missed a '$' prefix that should have been added. llvm-svn: 116690
2010-10-18 06:03:32 +08:00
#include <sys/stat.h>
#include <sys/sysctl.h>
#include <sys/types.h>
Fixed a race condition that could cause ProcessGDBRemote::DoResume() to return an error saying the resume timed out. Previously the thread that was trying to resume the process would eventually call ProcessGDBRemote::DoResume() which would broadcast an event over to the async GDB remote thread which would sent the continue packet to the remote gdb server. Right after this was sent, it would set a predicate boolean value (protected by a mutex and condition) and then the thread that issued the ProcessGDBRemote::DoResume() would then wait for that condition variable to be set. If the async gdb thread was too quick though, the predicate boolean value could have been set to true and back to false by the time the thread that issued the ProcessGDBRemote::DoResume() checks the boolean value. So we can't use the predicate value as a handshake. I have changed the code over to using a Event by having the GDB remote communication object post an event: GDBRemoteCommunication::eBroadcastBitRunPacketSent This allows reliable handshaking between the two threads and avoids the erroneous ProcessGDBRemote::DoResume() errors. Added a host backtrace service to allow in process backtraces when trying to track down tricky issues. I need to see if LLVM has any backtracing abilities abstracted in it already, and if so, use that, but I needed something ASAP for the current issue I was working on. The static function is: void Host::Backtrace (Stream &strm, uint32_t max_frames); And it will backtrace at most "max_frames" frames for the current thread and can be used with any of the Stream subclasses for logging. llvm-svn: 120793
2010-12-03 14:02:24 +08:00
#include <unistd.h>
Added a new Host call to find LLDB related paths: static bool Host::GetLLDBPath (lldb::PathType path_type, FileSpec &file_spec); This will fill in "file_spec" with an appropriate path that is appropriate for the current Host OS. MacOSX will return paths within the LLDB.framework, and other unixes will return the paths they want. The current PathType enums are: typedef enum PathType { ePathTypeLLDBShlibDir, // The directory where the lldb.so (unix) or LLDB mach-o file in LLDB.framework (MacOSX) exists ePathTypeSupportExecutableDir, // Find LLDB support executable directory (debugserver, etc) ePathTypeHeaderDir, // Find LLDB header file directory ePathTypePythonDir // Find Python modules (PYTHONPATH) directory } PathType; All places that were finding executables are and python paths are now updated to use this Host call. Added another new host call to launch the inferior in a terminal. This ability will be very host specific and doesn't need to be supported on all systems. MacOSX currently will create a new .command file and tell Terminal.app to open the .command file. It also uses the new "darwin-debug" app which is a small app that uses posix to exec (no fork) and stop at the entry point of the program. The GDB remote plug-in is almost able launch a process and attach to it, it currently will spawn the process, but it won't attach to it just yet. This will let LLDB not have to share the terminal with another process and a new terminal window will pop up when you launch. This won't get hooked up until we work out all of the kinks. The new Host function is: static lldb::pid_t Host::LaunchInNewTerminal ( const char **argv, // argv[0] is executable const char **envp, const ArchSpec *arch_spec, bool stop_at_entry, bool disable_aslr); Cleaned up FileSpec::GetPath to not use strncpy() as it was always zero filling the entire path buffer. Fixed an issue with the dynamic checker function where I missed a '$' prefix that should have been added. llvm-svn: 116690
2010-10-18 06:03:32 +08:00
#include "lldb/Core/ArchSpec.h"
#include "lldb/Core/Communication.h"
#include "lldb/Core/DataBufferHeap.h"
#include "lldb/Core/DataExtractor.h"
#include "lldb/Core/Log.h"
Moved the execution context that was in the Debugger into the CommandInterpreter where it was always being used. Make sure that Modules can track their object file offsets correctly to allow opening of sub object files (like the "__commpage" on darwin). Modified the Platforms to be able to launch processes. The first part of this move is the platform soon will become the entity that launches your program and when it does, it uses a new ProcessLaunchInfo class which encapsulates all process launching settings. This simplifies the internal APIs needed for launching. I want to slowly phase out process launching from the process classes, so for now we can still launch just as we used to, but eventually the platform is the object that should do the launching. Modified the Host::LaunchProcess in the MacOSX Host.mm to correctly be able to launch processes with all of the new eLaunchFlag settings. Modified any code that was manually launching processes to use the Host::LaunchProcess functions. Fixed an issue where lldb_private::Args had implicitly defined copy constructors that could do the wrong thing. This has now been fixed by adding an appropriate copy constructor and assignment operator. Make sure we don't add empty ModuleSP entries to a module list. Fixed the commpage module creation on MacOSX, but we still need to train the MacOSX dynamic loader to not get rid of it when it doesn't have an entry in the all image infos. Abstracted many more calls from in ProcessGDBRemote down into the GDBRemoteCommunicationClient subclass to make the classes cleaner and more efficient. Fixed the default iOS ARM register context to be correct and also added support for targets that don't support the qThreadStopInfo packet by selecting the current thread (only if needed) and then sending a stop reply packet. Debugserver can now start up with a --unix-socket (-u for short) and can then bind to port zero and send the port it bound to to a listening process on the other end. This allows the GDB remote platform to spawn new GDB server instances (debugserver) to allow platform debugging. llvm-svn: 129351
2011-04-12 13:54:46 +08:00
#include "lldb/Core/Module.h"
#include "lldb/Core/ModuleSpec.h"
Added a new Host call to find LLDB related paths: static bool Host::GetLLDBPath (lldb::PathType path_type, FileSpec &file_spec); This will fill in "file_spec" with an appropriate path that is appropriate for the current Host OS. MacOSX will return paths within the LLDB.framework, and other unixes will return the paths they want. The current PathType enums are: typedef enum PathType { ePathTypeLLDBShlibDir, // The directory where the lldb.so (unix) or LLDB mach-o file in LLDB.framework (MacOSX) exists ePathTypeSupportExecutableDir, // Find LLDB support executable directory (debugserver, etc) ePathTypeHeaderDir, // Find LLDB header file directory ePathTypePythonDir // Find Python modules (PYTHONPATH) directory } PathType; All places that were finding executables are and python paths are now updated to use this Host call. Added another new host call to launch the inferior in a terminal. This ability will be very host specific and doesn't need to be supported on all systems. MacOSX currently will create a new .command file and tell Terminal.app to open the .command file. It also uses the new "darwin-debug" app which is a small app that uses posix to exec (no fork) and stop at the entry point of the program. The GDB remote plug-in is almost able launch a process and attach to it, it currently will spawn the process, but it won't attach to it just yet. This will let LLDB not have to share the terminal with another process and a new terminal window will pop up when you launch. This won't get hooked up until we work out all of the kinks. The new Host function is: static lldb::pid_t Host::LaunchInNewTerminal ( const char **argv, // argv[0] is executable const char **envp, const ArchSpec *arch_spec, bool stop_at_entry, bool disable_aslr); Cleaned up FileSpec::GetPath to not use strncpy() as it was always zero filling the entire path buffer. Fixed an issue with the dynamic checker function where I missed a '$' prefix that should have been added. llvm-svn: 116690
2010-10-18 06:03:32 +08:00
#include "lldb/Core/StreamFile.h"
#include "lldb/Core/StreamString.h"
#include "lldb/Core/StructuredData.h"
#include "lldb/Host/ConnectionFileDescriptor.h"
Moved the execution context that was in the Debugger into the CommandInterpreter where it was always being used. Make sure that Modules can track their object file offsets correctly to allow opening of sub object files (like the "__commpage" on darwin). Modified the Platforms to be able to launch processes. The first part of this move is the platform soon will become the entity that launches your program and when it does, it uses a new ProcessLaunchInfo class which encapsulates all process launching settings. This simplifies the internal APIs needed for launching. I want to slowly phase out process launching from the process classes, so for now we can still launch just as we used to, but eventually the platform is the object that should do the launching. Modified the Host::LaunchProcess in the MacOSX Host.mm to correctly be able to launch processes with all of the new eLaunchFlag settings. Modified any code that was manually launching processes to use the Host::LaunchProcess functions. Fixed an issue where lldb_private::Args had implicitly defined copy constructors that could do the wrong thing. This has now been fixed by adding an appropriate copy constructor and assignment operator. Make sure we don't add empty ModuleSP entries to a module list. Fixed the commpage module creation on MacOSX, but we still need to train the MacOSX dynamic loader to not get rid of it when it doesn't have an entry in the all image infos. Abstracted many more calls from in ProcessGDBRemote down into the GDBRemoteCommunicationClient subclass to make the classes cleaner and more efficient. Fixed the default iOS ARM register context to be correct and also added support for targets that don't support the qThreadStopInfo packet by selecting the current thread (only if needed) and then sending a stop reply packet. Debugserver can now start up with a --unix-socket (-u for short) and can then bind to port zero and send the port it bound to to a listening process on the other end. This allows the GDB remote platform to spawn new GDB server instances (debugserver) to allow platform debugging. llvm-svn: 129351
2011-04-12 13:54:46 +08:00
#include "lldb/Host/Endian.h"
#include "lldb/Host/FileSpec.h"
#include "lldb/Host/FileSystem.h"
#include "lldb/Host/HostInfo.h"
#include "lldb/Host/ThreadLauncher.h"
Moved the execution context that was in the Debugger into the CommandInterpreter where it was always being used. Make sure that Modules can track their object file offsets correctly to allow opening of sub object files (like the "__commpage" on darwin). Modified the Platforms to be able to launch processes. The first part of this move is the platform soon will become the entity that launches your program and when it does, it uses a new ProcessLaunchInfo class which encapsulates all process launching settings. This simplifies the internal APIs needed for launching. I want to slowly phase out process launching from the process classes, so for now we can still launch just as we used to, but eventually the platform is the object that should do the launching. Modified the Host::LaunchProcess in the MacOSX Host.mm to correctly be able to launch processes with all of the new eLaunchFlag settings. Modified any code that was manually launching processes to use the Host::LaunchProcess functions. Fixed an issue where lldb_private::Args had implicitly defined copy constructors that could do the wrong thing. This has now been fixed by adding an appropriate copy constructor and assignment operator. Make sure we don't add empty ModuleSP entries to a module list. Fixed the commpage module creation on MacOSX, but we still need to train the MacOSX dynamic loader to not get rid of it when it doesn't have an entry in the all image infos. Abstracted many more calls from in ProcessGDBRemote down into the GDBRemoteCommunicationClient subclass to make the classes cleaner and more efficient. Fixed the default iOS ARM register context to be correct and also added support for targets that don't support the qThreadStopInfo packet by selecting the current thread (only if needed) and then sending a stop reply packet. Debugserver can now start up with a --unix-socket (-u for short) and can then bind to port zero and send the port it bound to to a listening process on the other end. This allows the GDB remote platform to spawn new GDB server instances (debugserver) to allow platform debugging. llvm-svn: 129351
2011-04-12 13:54:46 +08:00
#include "lldb/Target/Platform.h"
#include "lldb/Target/Process.h"
Moved the execution context that was in the Debugger into the CommandInterpreter where it was always being used. Make sure that Modules can track their object file offsets correctly to allow opening of sub object files (like the "__commpage" on darwin). Modified the Platforms to be able to launch processes. The first part of this move is the platform soon will become the entity that launches your program and when it does, it uses a new ProcessLaunchInfo class which encapsulates all process launching settings. This simplifies the internal APIs needed for launching. I want to slowly phase out process launching from the process classes, so for now we can still launch just as we used to, but eventually the platform is the object that should do the launching. Modified the Host::LaunchProcess in the MacOSX Host.mm to correctly be able to launch processes with all of the new eLaunchFlag settings. Modified any code that was manually launching processes to use the Host::LaunchProcess functions. Fixed an issue where lldb_private::Args had implicitly defined copy constructors that could do the wrong thing. This has now been fixed by adding an appropriate copy constructor and assignment operator. Make sure we don't add empty ModuleSP entries to a module list. Fixed the commpage module creation on MacOSX, but we still need to train the MacOSX dynamic loader to not get rid of it when it doesn't have an entry in the all image infos. Abstracted many more calls from in ProcessGDBRemote down into the GDBRemoteCommunicationClient subclass to make the classes cleaner and more efficient. Fixed the default iOS ARM register context to be correct and also added support for targets that don't support the qThreadStopInfo packet by selecting the current thread (only if needed) and then sending a stop reply packet. Debugserver can now start up with a --unix-socket (-u for short) and can then bind to port zero and send the port it bound to to a listening process on the other end. This allows the GDB remote platform to spawn new GDB server instances (debugserver) to allow platform debugging. llvm-svn: 129351
2011-04-12 13:54:46 +08:00
#include "lldb/Utility/CleanUp.h"
#include "lldb/Utility/NameMatches.h"
#include "cfcpp/CFCBundle.h"
Added a new Host call to find LLDB related paths: static bool Host::GetLLDBPath (lldb::PathType path_type, FileSpec &file_spec); This will fill in "file_spec" with an appropriate path that is appropriate for the current Host OS. MacOSX will return paths within the LLDB.framework, and other unixes will return the paths they want. The current PathType enums are: typedef enum PathType { ePathTypeLLDBShlibDir, // The directory where the lldb.so (unix) or LLDB mach-o file in LLDB.framework (MacOSX) exists ePathTypeSupportExecutableDir, // Find LLDB support executable directory (debugserver, etc) ePathTypeHeaderDir, // Find LLDB header file directory ePathTypePythonDir // Find Python modules (PYTHONPATH) directory } PathType; All places that were finding executables are and python paths are now updated to use this Host call. Added another new host call to launch the inferior in a terminal. This ability will be very host specific and doesn't need to be supported on all systems. MacOSX currently will create a new .command file and tell Terminal.app to open the .command file. It also uses the new "darwin-debug" app which is a small app that uses posix to exec (no fork) and stop at the entry point of the program. The GDB remote plug-in is almost able launch a process and attach to it, it currently will spawn the process, but it won't attach to it just yet. This will let LLDB not have to share the terminal with another process and a new terminal window will pop up when you launch. This won't get hooked up until we work out all of the kinks. The new Host function is: static lldb::pid_t Host::LaunchInNewTerminal ( const char **argv, // argv[0] is executable const char **envp, const ArchSpec *arch_spec, bool stop_at_entry, bool disable_aslr); Cleaned up FileSpec::GetPath to not use strncpy() as it was always zero filling the entire path buffer. Fixed an issue with the dynamic checker function where I missed a '$' prefix that should have been added. llvm-svn: 116690
2010-10-18 06:03:32 +08:00
#include "cfcpp/CFCMutableArray.h"
#include "cfcpp/CFCMutableDictionary.h"
#include "cfcpp/CFCReleaser.h"
#include "cfcpp/CFCString.h"
#include <objc/objc-auto.h>
#include <CoreFoundation/CoreFoundation.h>
#include <Foundation/Foundation.h>
Moved the execution context that was in the Debugger into the CommandInterpreter where it was always being used. Make sure that Modules can track their object file offsets correctly to allow opening of sub object files (like the "__commpage" on darwin). Modified the Platforms to be able to launch processes. The first part of this move is the platform soon will become the entity that launches your program and when it does, it uses a new ProcessLaunchInfo class which encapsulates all process launching settings. This simplifies the internal APIs needed for launching. I want to slowly phase out process launching from the process classes, so for now we can still launch just as we used to, but eventually the platform is the object that should do the launching. Modified the Host::LaunchProcess in the MacOSX Host.mm to correctly be able to launch processes with all of the new eLaunchFlag settings. Modified any code that was manually launching processes to use the Host::LaunchProcess functions. Fixed an issue where lldb_private::Args had implicitly defined copy constructors that could do the wrong thing. This has now been fixed by adding an appropriate copy constructor and assignment operator. Make sure we don't add empty ModuleSP entries to a module list. Fixed the commpage module creation on MacOSX, but we still need to train the MacOSX dynamic loader to not get rid of it when it doesn't have an entry in the all image infos. Abstracted many more calls from in ProcessGDBRemote down into the GDBRemoteCommunicationClient subclass to make the classes cleaner and more efficient. Fixed the default iOS ARM register context to be correct and also added support for targets that don't support the qThreadStopInfo packet by selecting the current thread (only if needed) and then sending a stop reply packet. Debugserver can now start up with a --unix-socket (-u for short) and can then bind to port zero and send the port it bound to to a listening process on the other end. This allows the GDB remote platform to spawn new GDB server instances (debugserver) to allow platform debugging. llvm-svn: 129351
2011-04-12 13:54:46 +08:00
#ifndef _POSIX_SPAWN_DISABLE_ASLR
#define _POSIX_SPAWN_DISABLE_ASLR 0x0100
Moved the execution context that was in the Debugger into the CommandInterpreter where it was always being used. Make sure that Modules can track their object file offsets correctly to allow opening of sub object files (like the "__commpage" on darwin). Modified the Platforms to be able to launch processes. The first part of this move is the platform soon will become the entity that launches your program and when it does, it uses a new ProcessLaunchInfo class which encapsulates all process launching settings. This simplifies the internal APIs needed for launching. I want to slowly phase out process launching from the process classes, so for now we can still launch just as we used to, but eventually the platform is the object that should do the launching. Modified the Host::LaunchProcess in the MacOSX Host.mm to correctly be able to launch processes with all of the new eLaunchFlag settings. Modified any code that was manually launching processes to use the Host::LaunchProcess functions. Fixed an issue where lldb_private::Args had implicitly defined copy constructors that could do the wrong thing. This has now been fixed by adding an appropriate copy constructor and assignment operator. Make sure we don't add empty ModuleSP entries to a module list. Fixed the commpage module creation on MacOSX, but we still need to train the MacOSX dynamic loader to not get rid of it when it doesn't have an entry in the all image infos. Abstracted many more calls from in ProcessGDBRemote down into the GDBRemoteCommunicationClient subclass to make the classes cleaner and more efficient. Fixed the default iOS ARM register context to be correct and also added support for targets that don't support the qThreadStopInfo packet by selecting the current thread (only if needed) and then sending a stop reply packet. Debugserver can now start up with a --unix-socket (-u for short) and can then bind to port zero and send the port it bound to to a listening process on the other end. This allows the GDB remote platform to spawn new GDB server instances (debugserver) to allow platform debugging. llvm-svn: 129351
2011-04-12 13:54:46 +08:00
#endif
extern "C" {
int __pthread_chdir(const char *path);
int __pthread_fchdir(int fildes);
}
using namespace lldb;
using namespace lldb_private;
bool Host::GetBundleDirectory(const FileSpec &file,
FileSpec &bundle_directory) {
#if defined(__APPLE__)
if (file.GetFileType() == FileSpec::eFileTypeDirectory) {
char path[PATH_MAX];
if (file.GetPath(path, sizeof(path))) {
CFCBundle bundle(path);
if (bundle.GetPath(path, sizeof(path))) {
bundle_directory.SetFile(path, false);
return true;
}
}
}
#endif
bundle_directory.Clear();
return false;
}
bool Host::ResolveExecutableInBundle(FileSpec &file) {
#if defined(__APPLE__)
if (file.GetFileType() == FileSpec::eFileTypeDirectory) {
char path[PATH_MAX];
if (file.GetPath(path, sizeof(path))) {
CFCBundle bundle(path);
CFCReleaser<CFURLRef> url(bundle.CopyExecutableURL());
if (url.get()) {
if (::CFURLGetFileSystemRepresentation(url.get(), YES, (UInt8 *)path,
sizeof(path))) {
file.SetFile(path, false);
return true;
}
}
}
}
#endif
return false;
}
static void *AcceptPIDFromInferior(void *arg) {
const char *connect_url = (const char *)arg;
ConnectionFileDescriptor file_conn;
Error error;
if (file_conn.Connect(connect_url, &error) == eConnectionStatusSuccess) {
char pid_str[256];
::memset(pid_str, 0, sizeof(pid_str));
ConnectionStatus status;
const size_t pid_str_len =
file_conn.Read(pid_str, sizeof(pid_str), 0, status, NULL);
if (pid_str_len > 0) {
int pid = atoi(pid_str);
return (void *)(intptr_t)pid;
}
}
return NULL;
}
static bool WaitForProcessToSIGSTOP(const lldb::pid_t pid,
const int timeout_in_seconds) {
const int time_delta_usecs = 100000;
const int num_retries = timeout_in_seconds / time_delta_usecs;
for (int i = 0; i < num_retries; i++) {
struct proc_bsdinfo bsd_info;
int error = ::proc_pidinfo(pid, PROC_PIDTBSDINFO, (uint64_t)0, &bsd_info,
PROC_PIDTBSDINFO_SIZE);
switch (error) {
case EINVAL:
case ENOTSUP:
case ESRCH:
case EPERM:
return false;
default:
break;
case 0:
if (bsd_info.pbi_status == SSTOP)
return true;
}
::usleep(time_delta_usecs);
}
return false;
}
#if !defined(__arm__) && !defined(__arm64__) && !defined(__aarch64__)
// static lldb::pid_t
// LaunchInNewTerminalWithCommandFile
//(
// const char **argv,
// const char **envp,
// const char *working_dir,
// const ArchSpec *arch_spec,
// bool stop_at_entry,
// bool disable_aslr
//)
//{
// if (!argv || !argv[0])
// return LLDB_INVALID_PROCESS_ID;
//
// OSStatus error = 0;
//
// FileSpec program (argv[0], false);
//
//
// std::string unix_socket_name;
//
// char temp_file_path[PATH_MAX];
// const char *tmpdir = ::getenv ("TMPDIR");
// if (tmpdir == NULL)
// tmpdir = "/tmp/";
// ::snprintf (temp_file_path, sizeof(temp_file_path), "%s%s-XXXXXX", tmpdir,
// program.GetFilename().AsCString());
//
// if (::mktemp (temp_file_path) == NULL)
// return LLDB_INVALID_PROCESS_ID;
//
// unix_socket_name.assign (temp_file_path);
//
// ::strlcat (temp_file_path, ".command", sizeof (temp_file_path));
//
// StreamFile command_file;
// command_file.GetFile().Open (temp_file_path,
// File::eOpenOptionWrite |
// File::eOpenOptionCanCreate,
// lldb::eFilePermissionsDefault);
//
// if (!command_file.GetFile().IsValid())
// return LLDB_INVALID_PROCESS_ID;
//
// FileSpec darwin_debug_file_spec;
// if (!HostInfo::GetLLDBPath (ePathTypeSupportExecutableDir,
// darwin_debug_file_spec))
// return LLDB_INVALID_PROCESS_ID;
// darwin_debug_file_spec.GetFilename().SetCString("darwin-debug");
//
// if (!darwin_debug_file_spec.Exists())
// return LLDB_INVALID_PROCESS_ID;
//
// char launcher_path[PATH_MAX];
// darwin_debug_file_spec.GetPath(launcher_path, sizeof(launcher_path));
// command_file.Printf("\"%s\" ", launcher_path);
//
// command_file.Printf("--unix-socket=%s ", unix_socket_name.c_str());
//
// if (arch_spec && arch_spec->IsValid())
// {
// command_file.Printf("--arch=%s ", arch_spec->GetArchitectureName());
// }
//
// if (disable_aslr)
// {
// command_file.PutCString("--disable-aslr ");
// }
//
// command_file.PutCString("-- ");
//
// if (argv)
// {
// for (size_t i=0; argv[i] != NULL; ++i)
// {
// command_file.Printf("\"%s\" ", argv[i]);
// }
// }
// command_file.PutCString("\necho Process exited with status $?\n");
// command_file.GetFile().Close();
// if (::chmod (temp_file_path, S_IRWXU | S_IRWXG) != 0)
// return LLDB_INVALID_PROCESS_ID;
//
// CFCMutableDictionary cf_env_dict;
//
// const bool can_create = true;
// if (envp)
// {
// for (size_t i=0; envp[i] != NULL; ++i)
// {
// const char *env_entry = envp[i];
// const char *equal_pos = strchr(env_entry, '=');
// if (equal_pos)
// {
// std::string env_key (env_entry, equal_pos);
// std::string env_val (equal_pos + 1);
// CFCString cf_env_key (env_key.c_str(), kCFStringEncodingUTF8);
// CFCString cf_env_val (env_val.c_str(), kCFStringEncodingUTF8);
// cf_env_dict.AddValue (cf_env_key.get(), cf_env_val.get(),
// can_create);
// }
// }
// }
//
// LSApplicationParameters app_params;
// ::memset (&app_params, 0, sizeof (app_params));
// app_params.flags = kLSLaunchDontAddToRecents | kLSLaunchAsync;
// app_params.argv = NULL;
// app_params.environment = (CFDictionaryRef)cf_env_dict.get();
//
// CFCReleaser<CFURLRef> command_file_url
// (::CFURLCreateFromFileSystemRepresentation (NULL,
// (const UInt8 *)temp_file_path,
// strlen(temp_file_path),
// false));
//
// CFCMutableArray urls;
//
// // Terminal.app will open the ".command" file we have created
// // and run our process inside it which will wait at the entry point
// // for us to attach.
// urls.AppendValue(command_file_url.get());
//
//
// lldb::pid_t pid = LLDB_INVALID_PROCESS_ID;
//
// Error lldb_error;
// // Sleep and wait a bit for debugserver to start to listen...
// char connect_url[128];
// ::snprintf (connect_url, sizeof(connect_url), "unix-accept://%s",
// unix_socket_name.c_str());
//
// // Spawn a new thread to accept incoming connection on the connect_url
// // so we can grab the pid from the inferior
// lldb::thread_t accept_thread = Host::ThreadCreate
// (unix_socket_name.c_str(),
// AcceptPIDFromInferior,
// connect_url,
// &lldb_error);
//
// ProcessSerialNumber psn;
// error = LSOpenURLsWithRole(urls.get(), kLSRolesShell, NULL, &app_params,
// &psn, 1);
// if (error == noErr)
// {
// thread_result_t accept_thread_result = NULL;
// if (Host::ThreadJoin (accept_thread, &accept_thread_result,
// &lldb_error))
// {
// if (accept_thread_result)
// {
// pid = (intptr_t)accept_thread_result;
//
// // Wait for process to be stopped the entry point by watching
// // for the process status to be set to SSTOP which indicates
// it it
// // SIGSTOP'ed at the entry point
// WaitForProcessToSIGSTOP (pid, 5);
// }
// }
// }
// else
// {
// Host::ThreadCancel (accept_thread, &lldb_error);
// }
//
// return pid;
//}
const char *applscript_in_new_tty = "tell application \"Terminal\"\n"
" activate\n"
" do script \"%s\"\n"
"end tell\n";
const char *applscript_in_existing_tty = "\
set the_shell_script to \"%s\"\n\
tell application \"Terminal\"\n\
repeat with the_window in (get windows)\n\
repeat with the_tab in tabs of the_window\n\
set the_tty to tty in the_tab\n\
if the_tty contains \"%s\" then\n\
if the_tab is not busy then\n\
set selected of the_tab to true\n\
set frontmost of the_window to true\n\
do script the_shell_script in the_tab\n\
return\n\
end if\n\
end if\n\
end repeat\n\
end repeat\n\
do script the_shell_script\n\
end tell\n";
static Error
LaunchInNewTerminalWithAppleScript(const char *exe_path,
ProcessLaunchInfo &launch_info) {
Error error;
char unix_socket_name[PATH_MAX] = "/tmp/XXXXXX";
if (::mktemp(unix_socket_name) == NULL) {
error.SetErrorString("failed to make temporary path for a unix socket");
return error;
}
StreamString command;
FileSpec darwin_debug_file_spec;
if (!HostInfo::GetLLDBPath(ePathTypeSupportExecutableDir,
darwin_debug_file_spec)) {
error.SetErrorString("can't locate the 'darwin-debug' executable");
return error;
}
darwin_debug_file_spec.GetFilename().SetCString("darwin-debug");
if (!darwin_debug_file_spec.Exists()) {
error.SetErrorStringWithFormat(
"the 'darwin-debug' executable doesn't exists at '%s'",
darwin_debug_file_spec.GetPath().c_str());
return error;
}
char launcher_path[PATH_MAX];
darwin_debug_file_spec.GetPath(launcher_path, sizeof(launcher_path));
const ArchSpec &arch_spec = launch_info.GetArchitecture();
// Only set the architecture if it is valid and if it isn't Haswell (x86_64h).
if (arch_spec.IsValid() &&
arch_spec.GetCore() != ArchSpec::eCore_x86_64_x86_64h)
command.Printf("arch -arch %s ", arch_spec.GetArchitectureName());
command.Printf("'%s' --unix-socket=%s", launcher_path, unix_socket_name);
if (arch_spec.IsValid())
command.Printf(" --arch=%s", arch_spec.GetArchitectureName());
FileSpec working_dir{launch_info.GetWorkingDirectory()};
if (working_dir)
command.Printf(" --working-dir '%s'", working_dir.GetCString());
else {
char cwd[PATH_MAX];
if (getcwd(cwd, PATH_MAX))
command.Printf(" --working-dir '%s'", cwd);
}
if (launch_info.GetFlags().Test(eLaunchFlagDisableASLR))
command.PutCString(" --disable-aslr");
// We are launching on this host in a terminal. So compare the environment on
// the host
// to what is supplied in the launch_info. Any items that aren't in the host
// environment
// need to be sent to darwin-debug. If we send all environment entries, we
// might blow the
// max command line length, so we only send user modified entries.
const char **envp =
launch_info.GetEnvironmentEntries().GetConstArgumentVector();
StringList host_env;
const size_t host_env_count = Host::GetEnvironment(host_env);
if (envp && envp[0]) {
const char *env_entry;
for (size_t env_idx = 0; (env_entry = envp[env_idx]) != NULL; ++env_idx) {
bool add_entry = true;
for (size_t i = 0; i < host_env_count; ++i) {
const char *host_env_entry = host_env.GetStringAtIndex(i);
if (strcmp(env_entry, host_env_entry) == 0) {
add_entry = false;
break;
}
}
if (add_entry) {
command.Printf(" --env='%s'", env_entry);
}
}
}
command.PutCString(" -- ");
const char **argv = launch_info.GetArguments().GetConstArgumentVector();
if (argv) {
for (size_t i = 0; argv[i] != NULL; ++i) {
if (i == 0)
command.Printf(" '%s'", exe_path);
else
command.Printf(" '%s'", argv[i]);
}
} else {
command.Printf(" '%s'", exe_path);
}
command.PutCString(" ; echo Process exited with status $?");
if (launch_info.GetFlags().Test(lldb::eLaunchFlagCloseTTYOnExit))
command.PutCString(" ; exit");
StreamString applescript_source;
const char *tty_command = command.GetString().c_str();
// if (tty_name && tty_name[0])
// {
// applescript_source.Printf (applscript_in_existing_tty,
// tty_command,
// tty_name);
// }
// else
// {
applescript_source.Printf(applscript_in_new_tty, tty_command);
// }
const char *script_source = applescript_source.GetString().c_str();
// puts (script_source);
NSAppleScript *applescript = [[NSAppleScript alloc]
initWithSource:[NSString stringWithCString:script_source
encoding:NSUTF8StringEncoding]];
lldb::pid_t pid = LLDB_INVALID_PROCESS_ID;
Error lldb_error;
// Sleep and wait a bit for debugserver to start to listen...
ConnectionFileDescriptor file_conn;
char connect_url[128];
::snprintf(connect_url, sizeof(connect_url), "unix-accept://%s",
unix_socket_name);
// Spawn a new thread to accept incoming connection on the connect_url
// so we can grab the pid from the inferior. We have to do this because we
// are sending an AppleScript that will launch a process in Terminal.app,
// in a shell and the shell will fork/exec a couple of times before we get
// to the process that we wanted to launch. So when our process actually
// gets launched, we will handshake with it and get the process ID for it.
HostThread accept_thread = ThreadLauncher::LaunchThread(
unix_socket_name, AcceptPIDFromInferior, connect_url, &lldb_error);
[applescript executeAndReturnError:nil];
thread_result_t accept_thread_result = NULL;
lldb_error = accept_thread.Join(&accept_thread_result);
if (lldb_error.Success() && accept_thread_result) {
pid = (intptr_t)accept_thread_result;
// Wait for process to be stopped at the entry point by watching
// for the process status to be set to SSTOP which indicates it it
// SIGSTOP'ed at the entry point
WaitForProcessToSIGSTOP(pid, 5);
}
FileSystem::Unlink(FileSpec{unix_socket_name, false});
[applescript release];
if (pid != LLDB_INVALID_PROCESS_ID)
launch_info.SetProcessID(pid);
return error;
}
#endif // #if !defined(__arm__) && !defined(__arm64__) && !defined(__aarch64__)
// On MacOSX CrashReporter will display a string for each shared library if
// the shared library has an exported symbol named "__crashreporter_info__".
static std::mutex &GetCrashReporterMutex() {
static std::mutex g_mutex;
return g_mutex;
}
extern "C" {
const char *__crashreporter_info__ = NULL;
}
asm(".desc ___crashreporter_info__, 0x10");
void Host::SetCrashDescriptionWithFormat(const char *format, ...) {
static StreamString g_crash_description;
std::lock_guard<std::mutex> guard(GetCrashReporterMutex());
if (format) {
va_list args;
va_start(args, format);
g_crash_description.GetString().clear();
g_crash_description.PrintfVarArg(format, args);
va_end(args);
__crashreporter_info__ = g_crash_description.GetData();
} else {
__crashreporter_info__ = NULL;
}
}
void Host::SetCrashDescription(const char *cstr) {
std::lock_guard<std::mutex> guard(GetCrashReporterMutex());
static std::string g_crash_description;
if (cstr) {
g_crash_description.assign(cstr);
__crashreporter_info__ = g_crash_description.c_str();
} else {
__crashreporter_info__ = NULL;
}
}
bool Host::OpenFileInExternalEditor(const FileSpec &file_spec,
uint32_t line_no) {
#if defined(__arm__) || defined(__arm64__) || defined(__aarch64__)
return false;
#else
// We attach this to an 'odoc' event to specify a particular selection
typedef struct {
int16_t reserved0; // must be zero
int16_t fLineNumber;
int32_t fSelStart;
int32_t fSelEnd;
uint32_t reserved1; // must be zero
uint32_t reserved2; // must be zero
} BabelAESelInfo;
Log *log(lldb_private::GetLogIfAnyCategoriesSet(LIBLLDB_LOG_HOST));
char file_path[PATH_MAX];
file_spec.GetPath(file_path, PATH_MAX);
CFCString file_cfstr(file_path, kCFStringEncodingUTF8);
CFCReleaser<CFURLRef> file_URL(::CFURLCreateWithFileSystemPath(
NULL, file_cfstr.get(), kCFURLPOSIXPathStyle, false));
if (log)
log->Printf(
"Sending source file: \"%s\" and line: %d to external editor.\n",
file_path, line_no);
long error;
BabelAESelInfo file_and_line_info = {
0, // reserved0
(int16_t)(line_no - 1), // fLineNumber (zero based line number)
1, // fSelStart
1024, // fSelEnd
0, // reserved1
0 // reserved2
};
AEKeyDesc file_and_line_desc;
error = ::AECreateDesc(typeUTF8Text, &file_and_line_info,
sizeof(file_and_line_info),
&(file_and_line_desc.descContent));
if (error != noErr) {
if (log)
log->Printf("Error creating AEDesc: %ld.\n", error);
return false;
}
file_and_line_desc.descKey = keyAEPosition;
static std::string g_app_name;
static FSRef g_app_fsref;
LSApplicationParameters app_params;
::memset(&app_params, 0, sizeof(app_params));
app_params.flags =
kLSLaunchDefaults | kLSLaunchDontAddToRecents | kLSLaunchDontSwitch;
char *external_editor = ::getenv("LLDB_EXTERNAL_EDITOR");
if (external_editor) {
if (log)
log->Printf("Looking for external editor \"%s\".\n", external_editor);
if (g_app_name.empty() ||
strcmp(g_app_name.c_str(), external_editor) != 0) {
CFCString editor_name(external_editor, kCFStringEncodingUTF8);
error = ::LSFindApplicationForInfo(kLSUnknownCreator, NULL,
editor_name.get(), &g_app_fsref, NULL);
// If we found the app, then store away the name so we don't have to
// re-look it up.
if (error != noErr) {
if (log)
log->Printf(
"Could not find External Editor application, error: %ld.\n",
error);
return false;
}
}
app_params.application = &g_app_fsref;
}
ProcessSerialNumber psn;
CFCReleaser<CFArrayRef> file_array(
CFArrayCreate(NULL, (const void **)file_URL.ptr_address(false), 1, NULL));
error = ::LSOpenURLsWithRole(file_array.get(), kLSRolesAll,
&file_and_line_desc, &app_params, &psn, 1);
AEDisposeDesc(&(file_and_line_desc.descContent));
if (error != noErr) {
if (log)
log->Printf("LSOpenURLsWithRole failed, error: %ld.\n", error);
return false;
}
return true;
#endif // #if !defined(__arm__) && !defined(__arm64__) && !defined(__aarch64__)
}
Fixed a race condition that could cause ProcessGDBRemote::DoResume() to return an error saying the resume timed out. Previously the thread that was trying to resume the process would eventually call ProcessGDBRemote::DoResume() which would broadcast an event over to the async GDB remote thread which would sent the continue packet to the remote gdb server. Right after this was sent, it would set a predicate boolean value (protected by a mutex and condition) and then the thread that issued the ProcessGDBRemote::DoResume() would then wait for that condition variable to be set. If the async gdb thread was too quick though, the predicate boolean value could have been set to true and back to false by the time the thread that issued the ProcessGDBRemote::DoResume() checks the boolean value. So we can't use the predicate value as a handshake. I have changed the code over to using a Event by having the GDB remote communication object post an event: GDBRemoteCommunication::eBroadcastBitRunPacketSent This allows reliable handshaking between the two threads and avoids the erroneous ProcessGDBRemote::DoResume() errors. Added a host backtrace service to allow in process backtraces when trying to track down tricky issues. I need to see if LLVM has any backtracing abilities abstracted in it already, and if so, use that, but I needed something ASAP for the current issue I was working on. The static function is: void Host::Backtrace (Stream &strm, uint32_t max_frames); And it will backtrace at most "max_frames" frames for the current thread and can be used with any of the Stream subclasses for logging. llvm-svn: 120793
2010-12-03 14:02:24 +08:00
size_t Host::GetEnvironment(StringList &env) {
char **host_env = *_NSGetEnviron();
char *env_entry;
size_t i;
for (i = 0; (env_entry = host_env[i]) != NULL; ++i)
env.AppendString(env_entry);
return i;
}
Added more platform support. There are now some new commands: platform status -- gets status information for the selected platform platform create <platform-name> -- creates a new instance of a remote platform platform list -- list all available platforms platform select -- select a platform instance as the current platform (not working yet) When using "platform create" it will create a remote platform and make it the selected platform. For instances for iPhone OS debugging on Mac OS X one can do: (lldb) platform create remote-ios --sdk-version=4.0 Remote platform: iOS platform SDK version: 4.0 SDK path: "/Developer/Platforms/iPhoneOS.platform/DeviceSupport/4.0" Not connected to a remote device. (lldb) file ~/Documents/a.out Current executable set to '~/Documents/a.out' (armv6). (lldb) image list [ 0] /Volumes/work/gclayton/Documents/devb/attach/a.out [ 1] /Developer/Platforms/iPhoneOS.platform/DeviceSupport/4.0/Symbols/usr/lib/dyld [ 2] /Developer/Platforms/iPhoneOS.platform/DeviceSupport/4.0/Symbols/usr/lib/libSystem.B.dylib Note that this is all happening prior to running _or_ connecting to a remote platform. Once connected to a remote platform the OS version might change which means we will need to update our dependecies. Also once we run, we will need to match up the actualy binaries with the actualy UUID's to files in the SDK, or download and cache them locally. This is just the start of the remote platforms, but this modification is the first iteration in getting the platforms really doing something. llvm-svn: 127934
2011-03-19 09:12:21 +08:00
static bool GetMacOSXProcessCPUType(ProcessInstanceInfo &process_info) {
if (process_info.ProcessIDIsValid()) {
// Make a new mib to stay thread safe
int mib[CTL_MAXNAME] = {
0,
};
size_t mib_len = CTL_MAXNAME;
if (::sysctlnametomib("sysctl.proc_cputype", mib, &mib_len))
return false;
mib[mib_len] = process_info.GetProcessID();
mib_len++;
cpu_type_t cpu, sub = 0;
size_t len = sizeof(cpu);
if (::sysctl(mib, mib_len, &cpu, &len, 0, 0) == 0) {
switch (cpu) {
case CPU_TYPE_I386:
sub = CPU_SUBTYPE_I386_ALL;
break;
case CPU_TYPE_X86_64:
sub = CPU_SUBTYPE_X86_64_ALL;
break;
#if defined(CPU_TYPE_ARM64) && defined(CPU_SUBTYPE_ARM64_ALL)
case CPU_TYPE_ARM64:
sub = CPU_SUBTYPE_ARM64_ALL;
break;
#endif
case CPU_TYPE_ARM: {
// Note that we fetched the cpu type from the PROCESS but we can't get a
// cpusubtype of the
// process -- we can only get the host's cpu subtype.
uint32_t cpusubtype = 0;
len = sizeof(cpusubtype);
if (::sysctlbyname("hw.cpusubtype", &cpusubtype, &len, NULL, 0) == 0)
sub = cpusubtype;
bool host_cpu_is_64bit;
uint32_t is64bit_capable;
size_t is64bit_capable_len = sizeof(is64bit_capable);
if (sysctlbyname("hw.cpu64bit_capable", &is64bit_capable,
&is64bit_capable_len, NULL, 0) == 0)
host_cpu_is_64bit = true;
else
host_cpu_is_64bit = false;
// if the host is an armv8 device, its cpusubtype will be in
// CPU_SUBTYPE_ARM64 numbering
// and we need to rewrite it to a reasonable CPU_SUBTYPE_ARM value
// instead.
if (host_cpu_is_64bit) {
sub = CPU_SUBTYPE_ARM_V7;
}
} break;
default:
break;
}
process_info.GetArchitecture().SetArchitecture(eArchTypeMachO, cpu, sub);
return true;
}
}
process_info.GetArchitecture().Clear();
return false;
}
static bool GetMacOSXProcessArgs(const ProcessInstanceInfoMatch *match_info_ptr,
ProcessInstanceInfo &process_info) {
if (process_info.ProcessIDIsValid()) {
int proc_args_mib[3] = {CTL_KERN, KERN_PROCARGS2,
(int)process_info.GetProcessID()};
size_t arg_data_size = 0;
if (::sysctl(proc_args_mib, 3, nullptr, &arg_data_size, NULL, 0) ||
arg_data_size == 0)
arg_data_size = 8192;
// Add a few bytes to the calculated length, I know we need to add at least
// one byte
// to this number otherwise we get junk back, so add 128 just in case...
DataBufferHeap arg_data(arg_data_size + 128, 0);
arg_data_size = arg_data.GetByteSize();
if (::sysctl(proc_args_mib, 3, arg_data.GetBytes(), &arg_data_size, NULL,
0) == 0) {
DataExtractor data(arg_data.GetBytes(), arg_data_size,
endian::InlHostByteOrder(), sizeof(void *));
lldb::offset_t offset = 0;
uint32_t argc = data.GetU32(&offset);
llvm::Triple &triple = process_info.GetArchitecture().GetTriple();
const llvm::Triple::ArchType triple_arch = triple.getArch();
const bool check_for_ios_simulator =
(triple_arch == llvm::Triple::x86 ||
triple_arch == llvm::Triple::x86_64);
const char *cstr = data.GetCStr(&offset);
if (cstr) {
process_info.GetExecutableFile().SetFile(cstr, false);
if (match_info_ptr == NULL ||
NameMatches(
process_info.GetExecutableFile().GetFilename().GetCString(),
match_info_ptr->GetNameMatchType(),
match_info_ptr->GetProcessInfo().GetName())) {
// Skip NULLs
while (1) {
const uint8_t *p = data.PeekData(offset, 1);
if ((p == NULL) || (*p != '\0'))
break;
++offset;
}
// Now extract all arguments
Args &proc_args = process_info.GetArguments();
for (int i = 0; i < static_cast<int>(argc); ++i) {
cstr = data.GetCStr(&offset);
if (cstr)
proc_args.AppendArgument(cstr);
}
Args &proc_env = process_info.GetEnvironmentEntries();
while ((cstr = data.GetCStr(&offset))) {
if (cstr[0] == '\0')
break;
if (check_for_ios_simulator) {
if (strncmp(cstr, "SIMULATOR_UDID=", strlen("SIMULATOR_UDID=")) ==
0)
process_info.GetArchitecture().GetTriple().setOS(
llvm::Triple::IOS);
else
process_info.GetArchitecture().GetTriple().setOS(
llvm::Triple::MacOSX);
}
proc_env.AppendArgument(cstr);
}
return true;
}
}
}
}
return false;
}
static bool GetMacOSXProcessUserAndGroup(ProcessInstanceInfo &process_info) {
if (process_info.ProcessIDIsValid()) {
int mib[4];
mib[0] = CTL_KERN;
mib[1] = KERN_PROC;
mib[2] = KERN_PROC_PID;
mib[3] = process_info.GetProcessID();
struct kinfo_proc proc_kinfo;
size_t proc_kinfo_size = sizeof(struct kinfo_proc);
if (::sysctl(mib, 4, &proc_kinfo, &proc_kinfo_size, NULL, 0) == 0) {
if (proc_kinfo_size > 0) {
process_info.SetParentProcessID(proc_kinfo.kp_eproc.e_ppid);
process_info.SetUserID(proc_kinfo.kp_eproc.e_pcred.p_ruid);
process_info.SetGroupID(proc_kinfo.kp_eproc.e_pcred.p_rgid);
process_info.SetEffectiveUserID(proc_kinfo.kp_eproc.e_ucred.cr_uid);
if (proc_kinfo.kp_eproc.e_ucred.cr_ngroups > 0)
process_info.SetEffectiveGroupID(
proc_kinfo.kp_eproc.e_ucred.cr_groups[0]);
else
process_info.SetEffectiveGroupID(UINT32_MAX);
return true;
}
}
}
process_info.SetParentProcessID(LLDB_INVALID_PROCESS_ID);
process_info.SetUserID(UINT32_MAX);
process_info.SetGroupID(UINT32_MAX);
process_info.SetEffectiveUserID(UINT32_MAX);
process_info.SetEffectiveGroupID(UINT32_MAX);
return false;
}
uint32_t Host::FindProcesses(const ProcessInstanceInfoMatch &match_info,
ProcessInstanceInfoList &process_infos) {
std::vector<struct kinfo_proc> kinfos;
int mib[3] = {CTL_KERN, KERN_PROC, KERN_PROC_ALL};
size_t pid_data_size = 0;
if (::sysctl(mib, 4, NULL, &pid_data_size, NULL, 0) != 0)
return 0;
// Add a few extra in case a few more show up
const size_t estimated_pid_count =
(pid_data_size / sizeof(struct kinfo_proc)) + 10;
kinfos.resize(estimated_pid_count);
pid_data_size = kinfos.size() * sizeof(struct kinfo_proc);
if (::sysctl(mib, 4, &kinfos[0], &pid_data_size, NULL, 0) != 0)
return 0;
const size_t actual_pid_count = (pid_data_size / sizeof(struct kinfo_proc));
bool all_users = match_info.GetMatchAllUsers();
const lldb::pid_t our_pid = getpid();
const uid_t our_uid = getuid();
for (size_t i = 0; i < actual_pid_count; i++) {
const struct kinfo_proc &kinfo = kinfos[i];
bool kinfo_user_matches = false;
if (all_users)
kinfo_user_matches = true;
else
kinfo_user_matches = kinfo.kp_eproc.e_pcred.p_ruid == our_uid;
// Special case, if lldb is being run as root we can attach to anything.
if (our_uid == 0)
kinfo_user_matches = true;
if (kinfo_user_matches == false || // Make sure the user is acceptable
static_cast<lldb::pid_t>(kinfo.kp_proc.p_pid) ==
our_pid || // Skip this process
kinfo.kp_proc.p_pid == 0 || // Skip kernel (kernel pid is zero)
kinfo.kp_proc.p_stat == SZOMB || // Zombies are bad, they like brains...
kinfo.kp_proc.p_flag & P_TRACED || // Being debugged?
kinfo.kp_proc.p_flag & P_WEXIT || // Working on exiting?
kinfo.kp_proc.p_flag & P_TRANSLATED) // Skip translated ppc (Rosetta)
continue;
ProcessInstanceInfo process_info;
process_info.SetProcessID(kinfo.kp_proc.p_pid);
process_info.SetParentProcessID(kinfo.kp_eproc.e_ppid);
process_info.SetUserID(kinfo.kp_eproc.e_pcred.p_ruid);
process_info.SetGroupID(kinfo.kp_eproc.e_pcred.p_rgid);
process_info.SetEffectiveUserID(kinfo.kp_eproc.e_ucred.cr_uid);
if (kinfo.kp_eproc.e_ucred.cr_ngroups > 0)
process_info.SetEffectiveGroupID(kinfo.kp_eproc.e_ucred.cr_groups[0]);
else
process_info.SetEffectiveGroupID(UINT32_MAX);
// Make sure our info matches before we go fetch the name and cpu type
if (match_info.Matches(process_info)) {
// Get CPU type first so we can know to look for iOS simulator is we have
// x86 or x86_64
if (GetMacOSXProcessCPUType(process_info)) {
if (GetMacOSXProcessArgs(&match_info, process_info)) {
if (match_info.Matches(process_info))
process_infos.Append(process_info);
}
}
}
}
return process_infos.GetSize();
}
bool Host::GetProcessInfo(lldb::pid_t pid, ProcessInstanceInfo &process_info) {
process_info.SetProcessID(pid);
bool success = false;
// Get CPU type first so we can know to look for iOS simulator is we have x86
// or x86_64
if (GetMacOSXProcessCPUType(process_info))
success = true;
if (GetMacOSXProcessArgs(NULL, process_info))
success = true;
if (GetMacOSXProcessUserAndGroup(process_info))
success = true;
if (success)
return true;
process_info.Clear();
return false;
}
#if !NO_XPC_SERVICES
static void PackageXPCArguments(xpc_object_t message, const char *prefix,
const Args &args) {
size_t count = args.GetArgumentCount();
char buf[50]; // long enough for 'argXXX'
memset(buf, 0, 50);
sprintf(buf, "%sCount", prefix);
xpc_dictionary_set_int64(message, buf, count);
for (size_t i = 0; i < count; i++) {
memset(buf, 0, 50);
sprintf(buf, "%s%zi", prefix, i);
xpc_dictionary_set_string(message, buf, args.GetArgumentAtIndex(i));
}
}
Moved the execution context that was in the Debugger into the CommandInterpreter where it was always being used. Make sure that Modules can track their object file offsets correctly to allow opening of sub object files (like the "__commpage" on darwin). Modified the Platforms to be able to launch processes. The first part of this move is the platform soon will become the entity that launches your program and when it does, it uses a new ProcessLaunchInfo class which encapsulates all process launching settings. This simplifies the internal APIs needed for launching. I want to slowly phase out process launching from the process classes, so for now we can still launch just as we used to, but eventually the platform is the object that should do the launching. Modified the Host::LaunchProcess in the MacOSX Host.mm to correctly be able to launch processes with all of the new eLaunchFlag settings. Modified any code that was manually launching processes to use the Host::LaunchProcess functions. Fixed an issue where lldb_private::Args had implicitly defined copy constructors that could do the wrong thing. This has now been fixed by adding an appropriate copy constructor and assignment operator. Make sure we don't add empty ModuleSP entries to a module list. Fixed the commpage module creation on MacOSX, but we still need to train the MacOSX dynamic loader to not get rid of it when it doesn't have an entry in the all image infos. Abstracted many more calls from in ProcessGDBRemote down into the GDBRemoteCommunicationClient subclass to make the classes cleaner and more efficient. Fixed the default iOS ARM register context to be correct and also added support for targets that don't support the qThreadStopInfo packet by selecting the current thread (only if needed) and then sending a stop reply packet. Debugserver can now start up with a --unix-socket (-u for short) and can then bind to port zero and send the port it bound to to a listening process on the other end. This allows the GDB remote platform to spawn new GDB server instances (debugserver) to allow platform debugging. llvm-svn: 129351
2011-04-12 13:54:46 +08:00
/*
A valid authorizationRef means that
- there is the LaunchUsingXPCRightName rights in the /etc/authorization
- we have successfully copied the rights to be send over the XPC wire
Once obtained, it will be valid for as long as the process lives.
*/
static AuthorizationRef authorizationRef = NULL;
static Error getXPCAuthorization(ProcessLaunchInfo &launch_info) {
Error error;
Log *log(lldb_private::GetLogIfAllCategoriesSet(LIBLLDB_LOG_HOST |
LIBLLDB_LOG_PROCESS));
if ((launch_info.GetUserID() == 0) && !authorizationRef) {
OSStatus createStatus =
AuthorizationCreate(NULL, kAuthorizationEmptyEnvironment,
kAuthorizationFlagDefaults, &authorizationRef);
if (createStatus != errAuthorizationSuccess) {
error.SetError(1, eErrorTypeGeneric);
error.SetErrorString("Can't create authorizationRef.");
if (log) {
error.PutToLog(log, "%s", error.AsCString());
}
return error;
Moved the execution context that was in the Debugger into the CommandInterpreter where it was always being used. Make sure that Modules can track their object file offsets correctly to allow opening of sub object files (like the "__commpage" on darwin). Modified the Platforms to be able to launch processes. The first part of this move is the platform soon will become the entity that launches your program and when it does, it uses a new ProcessLaunchInfo class which encapsulates all process launching settings. This simplifies the internal APIs needed for launching. I want to slowly phase out process launching from the process classes, so for now we can still launch just as we used to, but eventually the platform is the object that should do the launching. Modified the Host::LaunchProcess in the MacOSX Host.mm to correctly be able to launch processes with all of the new eLaunchFlag settings. Modified any code that was manually launching processes to use the Host::LaunchProcess functions. Fixed an issue where lldb_private::Args had implicitly defined copy constructors that could do the wrong thing. This has now been fixed by adding an appropriate copy constructor and assignment operator. Make sure we don't add empty ModuleSP entries to a module list. Fixed the commpage module creation on MacOSX, but we still need to train the MacOSX dynamic loader to not get rid of it when it doesn't have an entry in the all image infos. Abstracted many more calls from in ProcessGDBRemote down into the GDBRemoteCommunicationClient subclass to make the classes cleaner and more efficient. Fixed the default iOS ARM register context to be correct and also added support for targets that don't support the qThreadStopInfo packet by selecting the current thread (only if needed) and then sending a stop reply packet. Debugserver can now start up with a --unix-socket (-u for short) and can then bind to port zero and send the port it bound to to a listening process on the other end. This allows the GDB remote platform to spawn new GDB server instances (debugserver) to allow platform debugging. llvm-svn: 129351
2011-04-12 13:54:46 +08:00
}
OSStatus rightsStatus =
AuthorizationRightGet(LaunchUsingXPCRightName, NULL);
if (rightsStatus != errAuthorizationSuccess) {
// No rights in the security database, Create it with the right prompt.
CFStringRef prompt =
CFSTR("Xcode is trying to take control of a root process.");
CFStringRef keys[] = {CFSTR("en")};
CFTypeRef values[] = {prompt};
CFDictionaryRef promptDict = CFDictionaryCreate(
kCFAllocatorDefault, (const void **)keys, (const void **)values, 1,
&kCFCopyStringDictionaryKeyCallBacks,
&kCFTypeDictionaryValueCallBacks);
CFStringRef keys1[] = {CFSTR("class"), CFSTR("group"), CFSTR("comment"),
CFSTR("default-prompt"), CFSTR("shared")};
CFTypeRef values1[] = {CFSTR("user"), CFSTR("admin"),
CFSTR(LaunchUsingXPCRightName), promptDict,
kCFBooleanFalse};
CFDictionaryRef dict = CFDictionaryCreate(
kCFAllocatorDefault, (const void **)keys1, (const void **)values1, 5,
&kCFCopyStringDictionaryKeyCallBacks,
&kCFTypeDictionaryValueCallBacks);
rightsStatus = AuthorizationRightSet(
authorizationRef, LaunchUsingXPCRightName, dict, NULL, NULL, NULL);
CFRelease(promptDict);
CFRelease(dict);
}
OSStatus copyRightStatus = errAuthorizationDenied;
if (rightsStatus == errAuthorizationSuccess) {
AuthorizationItem item1 = {LaunchUsingXPCRightName, 0, NULL, 0};
AuthorizationItem items[] = {item1};
AuthorizationRights requestedRights = {1, items};
AuthorizationFlags authorizationFlags =
kAuthorizationFlagInteractionAllowed | kAuthorizationFlagExtendRights;
copyRightStatus = AuthorizationCopyRights(
authorizationRef, &requestedRights, kAuthorizationEmptyEnvironment,
authorizationFlags, NULL);
}
if (copyRightStatus != errAuthorizationSuccess) {
// Eventually when the commandline supports running as root and the user
// is not
// logged in in the current audit session, we will need the trick in gdb
// where
// we ask the user to type in the root passwd in the terminal.
error.SetError(2, eErrorTypeGeneric);
error.SetErrorStringWithFormat(
"Launching as root needs root authorization.");
if (log) {
error.PutToLog(log, "%s", error.AsCString());
}
if (authorizationRef) {
AuthorizationFree(authorizationRef, kAuthorizationFlagDefaults);
authorizationRef = NULL;
}
}
}
return error;
}
#endif
static Error LaunchProcessXPC(const char *exe_path,
ProcessLaunchInfo &launch_info,
lldb::pid_t &pid) {
#if !NO_XPC_SERVICES
Error error = getXPCAuthorization(launch_info);
if (error.Fail())
return error;
Log *log(lldb_private::GetLogIfAllCategoriesSet(LIBLLDB_LOG_HOST |
LIBLLDB_LOG_PROCESS));
uid_t requested_uid = launch_info.GetUserID();
const char *xpc_service = nil;
bool send_auth = false;
AuthorizationExternalForm extForm;
if (requested_uid == 0) {
if (AuthorizationMakeExternalForm(authorizationRef, &extForm) ==
errAuthorizationSuccess) {
send_auth = true;
} else {
error.SetError(3, eErrorTypeGeneric);
error.SetErrorStringWithFormat("Launching root via XPC needs to "
"externalize authorization reference.");
if (log) {
error.PutToLog(log, "%s", error.AsCString());
}
return error;
}
xpc_service = LaunchUsingXPCRightName;
} else {
error.SetError(4, eErrorTypeGeneric);
error.SetErrorStringWithFormat(
"Launching via XPC is only currently available for root.");
if (log) {
error.PutToLog(log, "%s", error.AsCString());
}
return error;
}
xpc_connection_t conn = xpc_connection_create(xpc_service, NULL);
xpc_connection_set_event_handler(conn, ^(xpc_object_t event) {
xpc_type_t type = xpc_get_type(event);
if (type == XPC_TYPE_ERROR) {
if (event == XPC_ERROR_CONNECTION_INTERRUPTED) {
// The service has either canceled itself, crashed, or been terminated.
// The XPC connection is still valid and sending a message to it will
// re-launch the service.
// If the service is state-full, this is the time to initialize the new
// service.
return;
} else if (event == XPC_ERROR_CONNECTION_INVALID) {
// The service is invalid. Either the service name supplied to
// xpc_connection_create() is incorrect
// or we (this process) have canceled the service; we can do any cleanup
// of application state at this point.
// printf("Service disconnected");
return;
} else {
// printf("Unexpected error from service: %s",
// xpc_dictionary_get_string(event, XPC_ERROR_KEY_DESCRIPTION));
}
} else {
// printf("Received unexpected event in handler");
}
});
xpc_connection_set_finalizer_f(conn, xpc_finalizer_t(xpc_release));
xpc_connection_resume(conn);
xpc_object_t message = xpc_dictionary_create(nil, nil, 0);
if (send_auth) {
xpc_dictionary_set_data(message, LauncherXPCServiceAuthKey, extForm.bytes,
sizeof(AuthorizationExternalForm));
}
PackageXPCArguments(message, LauncherXPCServiceArgPrefxKey,
launch_info.GetArguments());
PackageXPCArguments(message, LauncherXPCServiceEnvPrefxKey,
launch_info.GetEnvironmentEntries());
// Posix spawn stuff.
xpc_dictionary_set_int64(message, LauncherXPCServiceCPUTypeKey,
launch_info.GetArchitecture().GetMachOCPUType());
xpc_dictionary_set_int64(message, LauncherXPCServicePosixspawnFlagsKey,
Host::GetPosixspawnFlags(launch_info));
const FileAction *file_action = launch_info.GetFileActionForFD(STDIN_FILENO);
if (file_action && file_action->GetPath()) {
xpc_dictionary_set_string(message, LauncherXPCServiceStdInPathKeyKey,
file_action->GetPath());
}
file_action = launch_info.GetFileActionForFD(STDOUT_FILENO);
if (file_action && file_action->GetPath()) {
xpc_dictionary_set_string(message, LauncherXPCServiceStdOutPathKeyKey,
file_action->GetPath());
}
file_action = launch_info.GetFileActionForFD(STDERR_FILENO);
if (file_action && file_action->GetPath()) {
xpc_dictionary_set_string(message, LauncherXPCServiceStdErrPathKeyKey,
file_action->GetPath());
}
xpc_object_t reply =
xpc_connection_send_message_with_reply_sync(conn, message);
xpc_type_t returnType = xpc_get_type(reply);
if (returnType == XPC_TYPE_DICTIONARY) {
pid = xpc_dictionary_get_int64(reply, LauncherXPCServiceChildPIDKey);
if (pid == 0) {
int errorType =
xpc_dictionary_get_int64(reply, LauncherXPCServiceErrorTypeKey);
int errorCode =
xpc_dictionary_get_int64(reply, LauncherXPCServiceCodeTypeKey);
error.SetError(errorCode, eErrorTypeGeneric);
error.SetErrorStringWithFormat(
"Problems with launching via XPC. Error type : %i, code : %i",
errorType, errorCode);
if (log) {
error.PutToLog(log, "%s", error.AsCString());
}
if (authorizationRef) {
AuthorizationFree(authorizationRef, kAuthorizationFlagDefaults);
authorizationRef = NULL;
}
}
} else if (returnType == XPC_TYPE_ERROR) {
error.SetError(5, eErrorTypeGeneric);
error.SetErrorStringWithFormat(
"Problems with launching via XPC. XPC error : %s",
xpc_dictionary_get_string(reply, XPC_ERROR_KEY_DESCRIPTION));
if (log) {
error.PutToLog(log, "%s", error.AsCString());
Moved the execution context that was in the Debugger into the CommandInterpreter where it was always being used. Make sure that Modules can track their object file offsets correctly to allow opening of sub object files (like the "__commpage" on darwin). Modified the Platforms to be able to launch processes. The first part of this move is the platform soon will become the entity that launches your program and when it does, it uses a new ProcessLaunchInfo class which encapsulates all process launching settings. This simplifies the internal APIs needed for launching. I want to slowly phase out process launching from the process classes, so for now we can still launch just as we used to, but eventually the platform is the object that should do the launching. Modified the Host::LaunchProcess in the MacOSX Host.mm to correctly be able to launch processes with all of the new eLaunchFlag settings. Modified any code that was manually launching processes to use the Host::LaunchProcess functions. Fixed an issue where lldb_private::Args had implicitly defined copy constructors that could do the wrong thing. This has now been fixed by adding an appropriate copy constructor and assignment operator. Make sure we don't add empty ModuleSP entries to a module list. Fixed the commpage module creation on MacOSX, but we still need to train the MacOSX dynamic loader to not get rid of it when it doesn't have an entry in the all image infos. Abstracted many more calls from in ProcessGDBRemote down into the GDBRemoteCommunicationClient subclass to make the classes cleaner and more efficient. Fixed the default iOS ARM register context to be correct and also added support for targets that don't support the qThreadStopInfo packet by selecting the current thread (only if needed) and then sending a stop reply packet. Debugserver can now start up with a --unix-socket (-u for short) and can then bind to port zero and send the port it bound to to a listening process on the other end. This allows the GDB remote platform to spawn new GDB server instances (debugserver) to allow platform debugging. llvm-svn: 129351
2011-04-12 13:54:46 +08:00
}
}
return error;
#else
Error error;
return error;
#endif
}
static bool ShouldLaunchUsingXPC(ProcessLaunchInfo &launch_info) {
bool result = false;
#if !NO_XPC_SERVICES
bool launchingAsRoot = launch_info.GetUserID() == 0;
bool currentUserIsRoot = HostInfo::GetEffectiveUserID() == 0;
if (launchingAsRoot && !currentUserIsRoot) {
// If current user is already root, we don't need XPC's help.
result = true;
}
#endif
return result;
}
Error Host::LaunchProcess(ProcessLaunchInfo &launch_info) {
Error error;
char exe_path[PATH_MAX];
PlatformSP host_platform_sp(Platform::GetHostPlatform());
ModuleSpec exe_module_spec(launch_info.GetExecutableFile(),
launch_info.GetArchitecture());
FileSpec::FileType file_type = exe_module_spec.GetFileSpec().GetFileType();
if (file_type != FileSpec::eFileTypeRegular) {
lldb::ModuleSP exe_module_sp;
error = host_platform_sp->ResolveExecutable(exe_module_spec, exe_module_sp,
NULL);
if (error.Fail())
return error;
if (exe_module_sp)
exe_module_spec.GetFileSpec() = exe_module_sp->GetFileSpec();
}
if (exe_module_spec.GetFileSpec().Exists()) {
exe_module_spec.GetFileSpec().GetPath(exe_path, sizeof(exe_path));
} else {
launch_info.GetExecutableFile().GetPath(exe_path, sizeof(exe_path));
error.SetErrorStringWithFormat("executable doesn't exist: '%s'", exe_path);
return error;
}
if (launch_info.GetFlags().Test(eLaunchFlagLaunchInTTY)) {
#if !defined(__arm__) && !defined(__arm64__) && !defined(__aarch64__)
return LaunchInNewTerminalWithAppleScript(exe_path, launch_info);
#else
error.SetErrorString("launching a process in a new terminal is not "
"supported on iOS devices");
return error;
#endif
}
lldb::pid_t pid = LLDB_INVALID_PROCESS_ID;
if (ShouldLaunchUsingXPC(launch_info)) {
error = LaunchProcessXPC(exe_path, launch_info, pid);
} else {
error = LaunchProcessPosixSpawn(exe_path, launch_info, pid);
}
if (pid != LLDB_INVALID_PROCESS_ID) {
// If all went well, then set the process ID into the launch info
launch_info.SetProcessID(pid);
// Make sure we reap any processes we spawn or we will have zombies.
if (!launch_info.MonitorProcess()) {
const bool monitor_signals = false;
Host::MonitorChildProcessCallback callback = nullptr;
if (!launch_info.GetFlags().Test(lldb::eLaunchFlagDontSetExitStatus))
callback = Process::SetProcessExitStatus;
StartMonitoringChildProcess(callback, pid, monitor_signals);
}
} else {
// Invalid process ID, something didn't go well
if (error.Success())
error.SetErrorString("process launch failed for unknown reasons");
}
return error;
}
Error Host::ShellExpandArguments(ProcessLaunchInfo &launch_info) {
Error error;
if (launch_info.GetFlags().Test(eLaunchFlagShellExpandArguments)) {
FileSpec expand_tool_spec;
if (!HostInfo::GetLLDBPath(lldb::ePathTypeSupportExecutableDir,
expand_tool_spec)) {
error.SetErrorString(
"could not get support executable directory for lldb-argdumper tool");
return error;
}
expand_tool_spec.AppendPathComponent("lldb-argdumper");
if (!expand_tool_spec.Exists()) {
error.SetErrorStringWithFormat(
"could not find the lldb-argdumper tool: %s",
expand_tool_spec.GetPath().c_str());
return error;
}
StreamString expand_tool_spec_stream;
expand_tool_spec_stream.Printf("\"%s\"",
expand_tool_spec.GetPath().c_str());
Args expand_command(expand_tool_spec_stream.GetData());
expand_command.AppendArguments(launch_info.GetArguments());
int status;
std::string output;
FileSpec cwd(launch_info.GetWorkingDirectory());
if (!cwd.Exists()) {
char *wd = getcwd(nullptr, 0);
if (wd == nullptr) {
error.SetErrorStringWithFormat(
"cwd does not exist; cannot launch with shell argument expansion");
return error;
} else {
FileSpec working_dir(wd, false);
free(wd);
launch_info.SetWorkingDirectory(working_dir);
}
}
RunShellCommand(expand_command, cwd, &status, nullptr, &output, 10);
if (status != 0) {
error.SetErrorStringWithFormat("lldb-argdumper exited with error %d",
status);
return error;
}
auto data_sp = StructuredData::ParseJSON(output);
if (!data_sp) {
error.SetErrorString("invalid JSON");
return error;
}
auto dict_sp = data_sp->GetAsDictionary();
if (!data_sp) {
error.SetErrorString("invalid JSON");
return error;
}
auto args_sp = dict_sp->GetObjectForDotSeparatedPath("arguments");
if (!args_sp) {
error.SetErrorString("invalid JSON");
return error;
Moved the execution context that was in the Debugger into the CommandInterpreter where it was always being used. Make sure that Modules can track their object file offsets correctly to allow opening of sub object files (like the "__commpage" on darwin). Modified the Platforms to be able to launch processes. The first part of this move is the platform soon will become the entity that launches your program and when it does, it uses a new ProcessLaunchInfo class which encapsulates all process launching settings. This simplifies the internal APIs needed for launching. I want to slowly phase out process launching from the process classes, so for now we can still launch just as we used to, but eventually the platform is the object that should do the launching. Modified the Host::LaunchProcess in the MacOSX Host.mm to correctly be able to launch processes with all of the new eLaunchFlag settings. Modified any code that was manually launching processes to use the Host::LaunchProcess functions. Fixed an issue where lldb_private::Args had implicitly defined copy constructors that could do the wrong thing. This has now been fixed by adding an appropriate copy constructor and assignment operator. Make sure we don't add empty ModuleSP entries to a module list. Fixed the commpage module creation on MacOSX, but we still need to train the MacOSX dynamic loader to not get rid of it when it doesn't have an entry in the all image infos. Abstracted many more calls from in ProcessGDBRemote down into the GDBRemoteCommunicationClient subclass to make the classes cleaner and more efficient. Fixed the default iOS ARM register context to be correct and also added support for targets that don't support the qThreadStopInfo packet by selecting the current thread (only if needed) and then sending a stop reply packet. Debugserver can now start up with a --unix-socket (-u for short) and can then bind to port zero and send the port it bound to to a listening process on the other end. This allows the GDB remote platform to spawn new GDB server instances (debugserver) to allow platform debugging. llvm-svn: 129351
2011-04-12 13:54:46 +08:00
}
auto args_array_sp = args_sp->GetAsArray();
if (!args_array_sp) {
error.SetErrorString("invalid JSON");
return error;
Moved the execution context that was in the Debugger into the CommandInterpreter where it was always being used. Make sure that Modules can track their object file offsets correctly to allow opening of sub object files (like the "__commpage" on darwin). Modified the Platforms to be able to launch processes. The first part of this move is the platform soon will become the entity that launches your program and when it does, it uses a new ProcessLaunchInfo class which encapsulates all process launching settings. This simplifies the internal APIs needed for launching. I want to slowly phase out process launching from the process classes, so for now we can still launch just as we used to, but eventually the platform is the object that should do the launching. Modified the Host::LaunchProcess in the MacOSX Host.mm to correctly be able to launch processes with all of the new eLaunchFlag settings. Modified any code that was manually launching processes to use the Host::LaunchProcess functions. Fixed an issue where lldb_private::Args had implicitly defined copy constructors that could do the wrong thing. This has now been fixed by adding an appropriate copy constructor and assignment operator. Make sure we don't add empty ModuleSP entries to a module list. Fixed the commpage module creation on MacOSX, but we still need to train the MacOSX dynamic loader to not get rid of it when it doesn't have an entry in the all image infos. Abstracted many more calls from in ProcessGDBRemote down into the GDBRemoteCommunicationClient subclass to make the classes cleaner and more efficient. Fixed the default iOS ARM register context to be correct and also added support for targets that don't support the qThreadStopInfo packet by selecting the current thread (only if needed) and then sending a stop reply packet. Debugserver can now start up with a --unix-socket (-u for short) and can then bind to port zero and send the port it bound to to a listening process on the other end. This allows the GDB remote platform to spawn new GDB server instances (debugserver) to allow platform debugging. llvm-svn: 129351
2011-04-12 13:54:46 +08:00
}
launch_info.GetArguments().Clear();
for (size_t i = 0; i < args_array_sp->GetSize(); i++) {
auto item_sp = args_array_sp->GetItemAtIndex(i);
if (!item_sp)
continue;
auto str_sp = item_sp->GetAsString();
if (!str_sp)
continue;
launch_info.GetArguments().AppendArgument(str_sp->GetValue().c_str());
}
}
return error;
}
HostThread Host::StartMonitoringChildProcess(
const Host::MonitorChildProcessCallback &callback, lldb::pid_t pid,
bool monitor_signals) {
unsigned long mask = DISPATCH_PROC_EXIT;
if (monitor_signals)
mask |= DISPATCH_PROC_SIGNAL;
Log *log(lldb_private::GetLogIfAnyCategoriesSet(LIBLLDB_LOG_HOST |
LIBLLDB_LOG_PROCESS));
dispatch_source_t source = ::dispatch_source_create(
DISPATCH_SOURCE_TYPE_PROC, pid, mask,
::dispatch_get_global_queue(DISPATCH_QUEUE_PRIORITY_DEFAULT, 0));
if (log)
log->Printf("Host::StartMonitoringChildProcess "
"(callback, pid=%i, monitor_signals=%i) "
"source = %p\n",
static_cast<int>(pid), monitor_signals,
reinterpret_cast<void *>(source));
if (source) {
Host::MonitorChildProcessCallback callback_copy = callback;
::dispatch_source_set_cancel_handler(source, ^{
dispatch_release(source);
});
::dispatch_source_set_event_handler(source, ^{
int status = 0;
int wait_pid = 0;
bool cancel = false;
bool exited = false;
do {
wait_pid = ::waitpid(pid, &status, 0);
} while (wait_pid < 0 && errno == EINTR);
if (wait_pid >= 0) {
int signal = 0;
int exit_status = 0;
const char *status_cstr = NULL;
if (WIFSTOPPED(status)) {
signal = WSTOPSIG(status);
status_cstr = "STOPPED";
} else if (WIFEXITED(status)) {
exit_status = WEXITSTATUS(status);
status_cstr = "EXITED";
exited = true;
} else if (WIFSIGNALED(status)) {
signal = WTERMSIG(status);
status_cstr = "SIGNALED";
exited = true;
exit_status = -1;
} else {
status_cstr = "???";
}
if (log)
log->Printf("::waitpid (pid = %llu, &status, 0) => pid = %i, status "
"= 0x%8.8x (%s), signal = %i, exit_status = %i",
pid, wait_pid, status, status_cstr, signal, exit_status);
if (callback_copy)
cancel = callback_copy(pid, exited, signal, exit_status);
if (exited || cancel) {
::dispatch_source_cancel(source);
}
}
});
::dispatch_resume(source);
}
return HostThread();
}
//----------------------------------------------------------------------
// Log to both stderr and to ASL Logging when running on MacOSX.
//----------------------------------------------------------------------
void Host::SystemLog(SystemLogType type, const char *format, va_list args) {
if (format && format[0]) {
static aslmsg g_aslmsg = NULL;
if (g_aslmsg == NULL) {
g_aslmsg = ::asl_new(ASL_TYPE_MSG);
char asl_key_sender[PATH_MAX];
snprintf(asl_key_sender, sizeof(asl_key_sender),
"com.apple.LLDB.framework");
::asl_set(g_aslmsg, ASL_KEY_SENDER, asl_key_sender);
}
// Copy the va_list so we can log this message twice
va_list copy_args;
va_copy(copy_args, args);
// Log to stderr
::vfprintf(stderr, format, copy_args);
va_end(copy_args);
int asl_level;
switch (type) {
case eSystemLogError:
asl_level = ASL_LEVEL_ERR;
break;
case eSystemLogWarning:
asl_level = ASL_LEVEL_WARNING;
break;
}
// Log to ASL
::asl_vlog(NULL, g_aslmsg, asl_level, format, args);
}
}
lldb::DataBufferSP Host::GetAuxvData(lldb_private::Process *process) {
return lldb::DataBufferSP();
}