[x86/Linux] paltest_pal_sxs_test1 PAL test fails

[seanshpark]

Current PAL test for x86/Linux gets one test fail.

exception_handling/pal_sxs/test1/paltest_pal_sxs_test1. Exit code: 139

PAL Test Results:
  Passed: 792
  Failed: 1

I get this log with the test program,

PAL_SXS test1 SIGSEGV handler (nil)
Starting pal_sxs test1 DllTest2
DLLTest2 PASSED
Starting pal_sxs test1 DllTest1
Segmentation fault (core dumped)

[seanshpark]

cc @parjong , @lucenticus

[janvorli]

I am not sure if it is related, but the code in ThrowExceptionFromContextInternal for x86 is using "push ebp" instead of "push_nonvol_reg ebp", so it is missing the .cfi annotation provided by that macro. I am also not sure if the cdecl calling convention of the ThrowExceptionFromContextInternal and ThrowExceptionHelper isn't possibly messing things up. Can you run the test under GDB, set breakpoint at ThrowExceptionHelper and when you hit it, check if bt debugger commands gets a stack trace that makes sense? You should hit the function twice (three times if everything worked), so check the stack trace in both cases.
The debugger will also break in on the sigsegv that we will be transforming to a C++ exception, so you'll need to do c when you hit it in the DllTest1 or 2. I hope GDB on ARM will be able to continue.
You can also set GDB to not to stop on the signal (handle SIGSEGV nostop noprint), which is probably better.

[seanshpark]

@janvorli , sure you know deep inside! thanks, I'll follow your guide.

[seanshpark]

This is the first caught on DllTest2

Program received signal SIGSEGV, Segmentation fault.
0xb7cea2ef in FailingFunction (p=0x22) at coreclr/src/pal/tests/palsuite/exception_handling/pal_sxs/test1/dlltest2.cpp:31
31	    *p = 1;          // Causes an access violation exception
(gdb) c
Continuing.

Breakpoint 1, ThrowExceptionHelper (ex=0xb78a2720) at coreclr/src/pal/src/exception/seh.cpp:231
231	    throw std::move(*ex);
(gdb) bt
#0  ThrowExceptionHelper (ex=0xb78a2720) at coreclr/src/pal/src/exception/seh.cpp:231
dotnet/coreclr#1  0xb7cea2ef in FailingFunction (p=0x22) at coreclr/src/pal/tests/palsuite/exception_handling/pal_sxs/test1/dlltest2.cpp:31
dotnet/coreclr#2  0xb7cea135 in DllTest2::$_1::operator() (this=0xbffff258, unused=0x0)
    at coreclr/src/pal/tests/palsuite/exception_handling/pal_sxs/test1/dlltest2.cpp:47
dotnet/coreclr#3  0xb7ce9db7 in DllTest2 () at coreclr/src/pal/tests/palsuite/exception_handling/pal_sxs/test1/dlltest2.cpp:52
dotnet/coreclr#4  0x08048dcb in main (argc=1, argv=0xbffff494)
    at coreclr/src/pal/tests/palsuite/exception_handling/pal_sxs/test1/exceptionsxs.cpp:98

and then, on the second one on the DllTest1

Program received signal SIGSEGV, Segmentation fault.
0xb7e633cb in FailingFunction (p=0x11) at coreclr/src/pal/tests/palsuite/exception_handling/pal_sxs/test1/dlltest1.cpp:32
32	    *p = 1;          // Causes an access violation exception
(gdb) c
Continuing.

Program received signal SIGSEGV, Segmentation fault.
0xbfffec78 in ?? ()
(gdb) bt
#0  0xbfffec78 in ?? ()
dotnet/coreclr#1  0xb7ffd000 in ?? ()
Backtrace stopped: previous frame inner to this frame (corrupt stack?)

[seanshpark]

While step through the codes with gdb,
In the second call from dlltest1, CatchHardwareExceptionHolder::IsEnabled() was returning false at seh.cpp:292 in function SEHProcessException().

        if (CatchHardwareExceptionHolder::IsEnabled())
        {
            PAL_ThrowExceptionFromContext(exception->GetContextRecord(), exception);
        }

It was that pThread value was different for dlltest1(0x80519e0) and dlltest2(0x80566c8) but SEHProcessException() accessed that of dlltest2(0x80566c8), so IsEnabled() was false, thus did not call PAL_ThrowExceptionFromContext function.

(gdb) p pThread
$2 = (CorUnix::CPalThread *) 0x80566c8
...
(gdb) p pThread
$4 = (CorUnix::CPalThread *) 0x80519e0

I maybe doing something wrong. I ran the program like this. I'll continue the search.

cd bin/obj/Linux.x86.Debug/src/pal/tests/palsuite
gdb exception_handling/pal_sxs/test1/paltest_pal_sxs_test1

[janvorli]

@seanshpark you are not doing anything wrong. Let me think about what can we wrong.

[janvorli]

@seanshpark can you please set breakpoints at the following three methods?
CatchHardwareExceptionHolder::IsEnabled
CatchHardwareExceptionHolder::CatchHardwareExceptionHolder
CatchHardwareExceptionHolder::~CatchHardwareExceptionHolder

Then run the test, when you hit the breakpoint step over the pThread assignment and dump the pThread value. The CatchHardwareExceptionHolder::CatchHardwareExceptionHolder increments m_hardwareExceptionHolderCount on the pThread, the CatchHardwareExceptionHolder::~CatchHardwareExceptionHolder decrements it. The IsEnabled is true if the count is not zero.
From this debugging, it should became clear why the count for the second thread was zero,

[seanshpark]

@janvorli , above two pThread values are from the CatchHardwareExceptionHolder::CatchHardwareExceptionHolder of DllTest2 and DllTest1 and they are different. But in CatchHardwareExceptionHolder::IsEnabled, pThread is that of DllTest2 which is 0x80566c8. So in DllTest1, m_hardwareExceptionHolderCount is zero so IsEnabled is false.
I'll try with AMD64 and see how that works.

[janvorli]

@seanshpark I have already tried with AMD64 yesterday, that's how I came up with the advice on how to debug it. If you record the sequence of calls to the three functions together with the pThread in each of it until it reaches the crash and paste it here, I should be able to tell where is the culprit.

[seanshpark]

@janvorli , thank you. this is the log.

(gdb) b src/pal/src/exception/seh.cpp:292
No source file named src/pal/src/exception/seh.cpp.
Make breakpoint pending on future shared library load? (y or [n]) y
Breakpoint 1 (src/pal/src/exception/seh.cpp:292) pending.
(gdb) b src/pal/src/exception/seh.cpp:364
No source file named src/pal/src/exception/seh.cpp.
Make breakpoint pending on future shared library load? (y or [n]) y
Breakpoint 2 (src/pal/src/exception/seh.cpp:364) pending.
(gdb) b src/pal/src/exception/seh.cpp:370
No source file named src/pal/src/exception/seh.cpp.
Make breakpoint pending on future shared library load? (y or [n]) y
Breakpoint 3 (src/pal/src/exception/seh.cpp:370) pending.
(gdb) b src/pal/src/thread/thread.cpp:172
No source file named src/pal/src/thread/thread.cpp.
Make breakpoint pending on future shared library load? (y or [n]) y
Breakpoint 4 (src/pal/src/thread/thread.cpp:172) pending.
(gdb) b src/pal/src/thread/thread.cpp:1967
No source file named src/pal/src/thread/thread.cpp.
Make breakpoint pending on future shared library load? (y or [n]) y
Breakpoint 5 (src/pal/src/thread/thread.cpp:1967) pending.
(gdb) r
Starting program: /home/maxwell/netcore/coreclr/bin/obj/Linux.x86.Debug/src/pal/tests/palsuite/exception_handling/pal_sxs/test1/paltest_pal_sxs_test1
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/i386-linux-gnu/libthread_db.so.1".
PAL_SXS test1 SIGSEGV handler (nil)

Breakpoint 5, CorUnix::CreateThreadObject (pThread=0x80519e0, pNewThread=0x80519e0, phThread=0xbfffee84) at /home/maxwell/netcore/coreclr/src/pal/src/thread/thread.cpp:1968
1968        *phThread = hThread;
(gdb) c
Continuing.

Breakpoint 5, CorUnix::CreateThreadObject (pThread=0x80566c8, pNewThread=0x80566c8, phThread=0xbfffee84) at /home/maxwell/netcore/coreclr/src/pal/src/thread/thread.cpp:1968
1968        *phThread = hThread;
(gdb) c
Continuing.
Starting pal_sxs test1 DllTest2

Breakpoint 2, CatchHardwareExceptionHolder::CatchHardwareExceptionHolder (this=0xbffff250) at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:364
364         CPalThread *pThread = InternalGetCurrentThread();
(gdb) n
365         ++pThread->m_hardwareExceptionHolderCount;
(gdb) p pThread
$1 = (CorUnix::CPalThread *) 0x80566c8
(gdb) c
Continuing.

Program received signal SIGSEGV, Segmentation fault.
0xb7cec1df in FailingFunction (p=0x22) at /home/maxwell/netcore/coreclr/src/pal/tests/palsuite/exception_handling/pal_sxs/test1/dlltest2.cpp:31
31          *p = 1;          // Causes an access violation exception
(gdb) c
Continuing.

Breakpoint 1, SEHProcessException (exception=0xbfffebb0) at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:292
292             if (CatchHardwareExceptionHolder::IsEnabled())
(gdb) s
CatchHardwareExceptionHolder::IsEnabled () at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:376
376         CPalThread *pThread = InternalGetCurrentThread();
(gdb) n
377         return pThread->IsHardwareExceptionsEnabled();
(gdb) p pThread
$2 = (CorUnix::CPalThread *) 0x80566c8
(gdb) c
Continuing.

Breakpoint 3, CatchHardwareExceptionHolder::~CatchHardwareExceptionHolder (this=0xbffff250) at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:370
370         CPalThread *pThread = InternalGetCurrentThread();
(gdb) n
371         --pThread->m_hardwareExceptionHolderCount;
(gdb) p pThread
$3 = (CorUnix::CPalThread *) 0x80566c8
(gdb) c
Continuing.
DLLTest2 PASSED
Starting pal_sxs test1 DllTest1

Breakpoint 2, CatchHardwareExceptionHolder::CatchHardwareExceptionHolder (this=0xbffff250) at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:364
364         CPalThread *pThread = InternalGetCurrentThread();
(gdb) n
365         ++pThread->m_hardwareExceptionHolderCount;
(gdb) p pThread
$4 = (CorUnix::CPalThread *) 0x80519e0
(gdb) c
Continuing.

Program received signal SIGSEGV, Segmentation fault.
0xb7e641df in FailingFunction (p=0x11) at /home/maxwell/netcore/coreclr/src/pal/tests/palsuite/exception_handling/pal_sxs/test1/dlltest1.cpp:31
31          *p = 1;          // Causes an access violation exception
(gdb) c
Continuing.

Breakpoint 1, SEHProcessException (exception=0xbfffebb0) at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:292
292             if (CatchHardwareExceptionHolder::IsEnabled())
(gdb) s
CatchHardwareExceptionHolder::IsEnabled () at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:376
376         CPalThread *pThread = InternalGetCurrentThread();
(gdb) n
377         return pThread->IsHardwareExceptionsEnabled();
(gdb) p pThread
$5 = (CorUnix::CPalThread *) 0x80566c8
(gdb) c
Continuing.

Program received signal SIGSEGV, Segmentation fault.
0xbfffecb8 in ?? ()
(gdb)

Compared to AMD64, for DllTest1, after checking that of DllTest2's pThread in CatchHardwareExceptionHolder::IsEnabled it checks again with that DllTest1, where x86 returns to somewhere wrong place, seems stack is corrupt.

I'm running with change push ebp with push_nonvol_reg ebp, with some addition to the macro;

+
+.macro push_nonvol_reg Register
+        push \Register
+        .cfi_adjust_cfa_offset 4
+        .cfi_rel_offset \Register, 0
+.endm
+
+.macro pop_nonvol_reg Register
+        pop \Register
+        .cfi_adjust_cfa_offset -4
+        .cfi_restore \Register
+.endm

(is 4 correct?)

if the cdecl calling convention of the ThrowExceptionFromContextInternal and ThrowExceptionHelper isn't possibly messing things up

I looked this but not sure how to look...

[janvorli]

@seanshpark one more thing. Can you repeat the same experiment, but also dump the current thread using gdb at each point you print the pThread? Just to see if the pThread and Linux thread match (in the sense that we would always get the same pair of pThread value and Linux thread id.
From your log, I suspect that the InternalGetCurrentThread doesn't work correctly.

[seanshpark]

@janvorli , thank you, right on it :)

[seanshpark]

(gdb) b src/pal/src/exception/seh.cpp:292
No source file named src/pal/src/exception/seh.cpp.
Make breakpoint pending on future shared library load? (y or [n]) y
Breakpoint 1 (src/pal/src/exception/seh.cpp:292) pending.
(gdb) b src/pal/src/exception/seh.cpp:364
No source file named src/pal/src/exception/seh.cpp.
Make breakpoint pending on future shared library load? (y or [n]) y
Breakpoint 2 (src/pal/src/exception/seh.cpp:364) pending.
(gdb) b src/pal/src/exception/seh.cpp:370
No source file named src/pal/src/exception/seh.cpp.
Make breakpoint pending on future shared library load? (y or [n]) y
Breakpoint 3 (src/pal/src/exception/seh.cpp:370) pending.
(gdb) b src/pal/src/thread/thread.cpp:172
No source file named src/pal/src/thread/thread.cpp.
Make breakpoint pending on future shared library load? (y or [n]) y
Breakpoint 4 (src/pal/src/thread/thread.cpp:172) pending.
(gdb) b src/pal/src/thread/thread.cpp:1967
No source file named src/pal/src/thread/thread.cpp.
Make breakpoint pending on future shared library load? (y or [n]) y
Breakpoint 5 (src/pal/src/thread/thread.cpp:1967) pending.
(gdb) r
Starting program: /home/maxwell/netcore/coreclr/bin/obj/Linux.x86.Debug/src/pal/tests/palsuite/exception_handling/pal_sxs/test1/paltest_pal_sxs_test1
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/i386-linux-gnu/libthread_db.so.1".
PAL_SXS test1 SIGSEGV handler (nil)

Breakpoint 5, CorUnix::CreateThreadObject (pThread=0x80519e0, pNewThread=0x80519e0, phThread=0xbfffee84)
    at /home/maxwell/netcore/coreclr/src/pal/src/thread/thread.cpp:1968
1968        *phThread = hThread;
(gdb) c
Continuing.

Breakpoint 5, CorUnix::CreateThreadObject (pThread=0x80566c8, pNewThread=0x80566c8, phThread=0xbfffee84)
    at /home/maxwell/netcore/coreclr/src/pal/src/thread/thread.cpp:1968
1968        *phThread = hThread;
(gdb) c
Continuing.
Starting pal_sxs test1 DllTest2

Breakpoint 2, CatchHardwareExceptionHolder::CatchHardwareExceptionHolder (this=0xbffff250)
    at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:364
364         CPalThread *pThread = InternalGetCurrentThread();
(gdb) n
365         ++pThread->m_hardwareExceptionHolderCount;
(gdb) p pThread
$1 = (CorUnix::CPalThread *) 0x80566c8
(gdb) p/x *pThread
$2 = {_vptr$CPalThread = 0xb7e508d0 <vtable for CorUnix::CPalThread+8>, m_pNext = 0x0, m_dwExitCode = 0x103,
  m_fExitCodeSet = 0x0, m_csLock = {DebugInfo = 0x8056d40, LockCount = 0x0, RecursionCount = 0x0, OwningThread = 0x0,
    LockSemaphore = 0x0, SpinCount = 0x0, bInternal = 0x1, dwInitState = 0x1, csnds = {rgNativeDataStorage = {
        0x0 <repeats 76 times>}, pvAlign = 0x0}}, m_fLockInitialized = 0x1, m_fIsDummy = 0x0, m_lRefCount = 0x1,
  m_pThreadObject = 0x805a420, m_threadId = 0x6fe0, m_dwLwpId = 0x0, m_pthreadSelf = 0xb78a4740,
  m_hardwareExceptionHolderCount = 0x0, m_lpStartAddress = 0x0, m_lpStartParameter = 0x0, m_bCreateSuspended = 0x0,
  m_iThreadPriority = 0x0, m_eThreadType = 0x0, m_startMutex = {__data = {__lock = 0x0, __count = 0x0, __owner = 0x0,
      __kind = 0x0, __nusers = 0x0, {__elision_data = {__espins = 0x0, __elision = 0x0}, __list = {__next = 0x0}}},
    __size = {0x0 <repeats 24 times>}, __align = 0x0}, m_startCond = {__data = {__lock = 0x0, __futex = 0x0,
      __total_seq = 0x0, __wakeup_seq = 0x0, __woken_seq = 0x0, __mutex = 0x0, __nwaiters = 0x0, __broadcast_seq = 0x0},
    __size = {0x0 <repeats 48 times>}, __align = 0x0}, m_fStartItemsInitialized = 0x1, m_fStartStatus = 0x0,
  m_fStartStatusSet = 0x0, m_stackBase = 0x0, m_stackLimit = 0x0, static s_dwDefaultThreadStackSize = 0x40000,
  m_fInPal = 0x1, synchronizationInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e507dc <vtable for CorUnix::CThreadSynchronizationInfo+8>},
    m_tsThreadState = 0x1, m_shridWaitAwakened = 0x70, m_lLocalSynchLockCount = {m_val = 0x0}, m_lSharedSynchLockCount = {
      m_val = 0x0}, m_leOwnedObjsList = {Flink = 0x80567e0, Blink = 0x80567e0}, m_ownedNamedMutexListLock = {
      DebugInfo = 0x8056d20, LockCount = 0x0, RecursionCount = 0x0, OwningThread = 0x0, LockSemaphore = 0x0,
      SpinCount = 0x0, bInternal = 0x0, dwInitState = 0x1, csnds = {rgNativeDataStorage = {0x0 <repeats 76 times>},
        pvAlign = 0x0}}, m_ownedNamedMutexListHead = 0x0, m_tnwdNativeData = {mutex = {__data = {__lock = 0x0,
          __count = 0x0, __owner = 0x0, __kind = 0x0, __nusers = 0x0, {__elision_data = {__espins = 0x0, __elision = 0x0},
            __list = {__next = 0x0}}}, __size = {0x0 <repeats 24 times>}, __align = 0x0}, cond = {__data = {__lock = 0x0,
          __futex = 0x0, __total_seq = 0x0, __wakeup_seq = 0x0, __woken_seq = 0x0, __mutex = 0x0, __nwaiters = 0x0,
          __broadcast_seq = 0x0}, __size = {0x0 <repeats 48 times>}, __align = 0x0}, iPred = 0x0, dwObjectIndex = 0x0,
      twrWakeupReason = 0x0, fInitialized = 0x1}, m_twiWaitInfo = {wtWaitType = 0x0, wdWaitDomain = 0x0, lObjCount = 0x0,
      lSharedObjCount = 0x0, pthrOwner = 0x80566c8, rgpWTLNodes = {0x0 <repeats 64 times>}},
    static PendingSignalingsArraySize = 0xa, m_lPendingSignalingCount = 0x0, m_rgpthrPendingSignalings = {0x0, 0x0, 0x0,
      0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0}, m_lePendingSignalingsOverflowList = {Flink = 0x80569f0, Blink = 0x80569f0}},
  suspensionInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e50948 <vtable for CorUnix::CThreadSuspensionInfo+8>}, m_fPending = 0x0,
---Type <return> to continue, or q <return> to quit---
    m_fSelfsusp = 0x0, m_fSuspendedForShutdown = 0x0, m_nBlockingPipe = 0xffffffff, m_lNumThreadsSuspendedByThisThread = {
      m_val = 0x0}, m_ptmSuspmutex = {__data = {__lock = 0x0, __count = 0x0, __owner = 0x0, __kind = 0x0, __nusers = 0x0, {
          __elision_data = {__espins = 0x0, __elision = 0x0}, __list = {__next = 0x0}}}, __size = {
        0x0 <repeats 24 times>}, __align = 0x0}, m_fSuspmutexInitialized = 0x1, m_semSusp = {__size = {
        0x0 <repeats 16 times>}, __align = 0x0}, m_semResume = {__size = {0x0 <repeats 16 times>}, __align = 0x0},
    m_fSemaphoresInitialized = 0x1}, sehInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e508e8 <vtable for CorUnix::CThreadSEHInfo+8>}, safe_state = 0x0,
    signal_code = 0x0}, tlsInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e50970 <vtable for CorUnix::CThreadTLSInfo+8>}, tlsSlots = {
      0x0 <repeats 64 times>}}, apcInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e50904 <vtable for CorUnix::CThreadApcInfo+8>}, m_ptainHead = 0x0,
    m_ptainTail = 0x0}, crtInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e50924 <vtable for CorUnix::CThreadCRTInfo+8>}, strtokContext = 0x0,
    wcstokContext = 0x0, localtimeBuffer = {tm_sec = 0x0, tm_min = 0x0, tm_hour = 0x0, tm_mday = 0x0, tm_mon = 0x0,
      tm_year = 0x0, tm_wday = 0x0, tm_yday = 0x0, tm_isdst = 0x0}, ctimeBuffer = {0x0 <repeats 26 times>}, ECVTBuffer = {
      0x0 <repeats 357 times>}}}
(gdb) c
Continuing.

Program received signal SIGSEGV, Segmentation fault.
0xb7cec1df in FailingFunction (p=0x22)
    at /home/maxwell/netcore/coreclr/src/pal/tests/palsuite/exception_handling/pal_sxs/test1/dlltest2.cpp:31
31          *p = 1;          // Causes an access violation exception
(gdb) c
Continuing.

Breakpoint 1, SEHProcessException (exception=0xbfffebb0) at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:292
292             if (CatchHardwareExceptionHolder::IsEnabled())
(gdb) s
CatchHardwareExceptionHolder::IsEnabled () at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:376
376         CPalThread *pThread = InternalGetCurrentThread();
(gdb) n
377         return pThread->IsHardwareExceptionsEnabled();
(gdb) p pThread
$3 = (CorUnix::CPalThread *) 0x80566c8
(gdb) p/x *pThread
$4 = {_vptr$CPalThread = 0xb7e508d0 <vtable for CorUnix::CPalThread+8>, m_pNext = 0x0, m_dwExitCode = 0x103,
  m_fExitCodeSet = 0x0, m_csLock = {DebugInfo = 0x8056d40, LockCount = 0x0, RecursionCount = 0x0, OwningThread = 0x0,
    LockSemaphore = 0x0, SpinCount = 0x0, bInternal = 0x1, dwInitState = 0x1, csnds = {rgNativeDataStorage = {
        0x0 <repeats 76 times>}, pvAlign = 0x0}}, m_fLockInitialized = 0x1, m_fIsDummy = 0x0, m_lRefCount = 0x1,
  m_pThreadObject = 0x805a420, m_threadId = 0x6fe0, m_dwLwpId = 0x0, m_pthreadSelf = 0xb78a4740,
  m_hardwareExceptionHolderCount = 0x1, m_lpStartAddress = 0x0, m_lpStartParameter = 0x0, m_bCreateSuspended = 0x0,
  m_iThreadPriority = 0x0, m_eThreadType = 0x0, m_startMutex = {__data = {__lock = 0x0, __count = 0x0, __owner = 0x0,
      __kind = 0x0, __nusers = 0x0, {__elision_data = {__espins = 0x0, __elision = 0x0}, __list = {__next = 0x0}}},
    __size = {0x0 <repeats 24 times>}, __align = 0x0}, m_startCond = {__data = {__lock = 0x0, __futex = 0x0,
      __total_seq = 0x0, __wakeup_seq = 0x0, __woken_seq = 0x0, __mutex = 0x0, __nwaiters = 0x0, __broadcast_seq = 0x0},
    __size = {0x0 <repeats 48 times>}, __align = 0x0}, m_fStartItemsInitialized = 0x1, m_fStartStatus = 0x0,
  m_fStartStatusSet = 0x0, m_stackBase = 0x0, m_stackLimit = 0x0, static s_dwDefaultThreadStackSize = 0x40000,
  m_fInPal = 0x1, synchronizationInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e507dc <vtable for CorUnix::CThreadSynchronizationInfo+8>},
    m_tsThreadState = 0x1, m_shridWaitAwakened = 0x70, m_lLocalSynchLockCount = {m_val = 0x0}, m_lSharedSynchLockCount = {
      m_val = 0x0}, m_leOwnedObjsList = {Flink = 0x80567e0, Blink = 0x80567e0}, m_ownedNamedMutexListLock = {
      DebugInfo = 0x8056d20, LockCount = 0x0, RecursionCount = 0x0, OwningThread = 0x0, LockSemaphore = 0x0,
      SpinCount = 0x0, bInternal = 0x0, dwInitState = 0x1, csnds = {rgNativeDataStorage = {0x0 <repeats 76 times>},
        pvAlign = 0x0}}, m_ownedNamedMutexListHead = 0x0, m_tnwdNativeData = {mutex = {__data = {__lock = 0x0,
          __count = 0x0, __owner = 0x0, __kind = 0x0, __nusers = 0x0, {__elision_data = {__espins = 0x0, __elision = 0x0},
            __list = {__next = 0x0}}}, __size = {0x0 <repeats 24 times>}, __align = 0x0}, cond = {__data = {__lock = 0x0,
          __futex = 0x0, __total_seq = 0x0, __wakeup_seq = 0x0, __woken_seq = 0x0, __mutex = 0x0, __nwaiters = 0x0,
          __broadcast_seq = 0x0}, __size = {0x0 <repeats 48 times>}, __align = 0x0}, iPred = 0x0, dwObjectIndex = 0x0,
      twrWakeupReason = 0x0, fInitialized = 0x1}, m_twiWaitInfo = {wtWaitType = 0x0, wdWaitDomain = 0x0, lObjCount = 0x0,
      lSharedObjCount = 0x0, pthrOwner = 0x80566c8, rgpWTLNodes = {0x0 <repeats 64 times>}},
    static PendingSignalingsArraySize = 0xa, m_lPendingSignalingCount = 0x0, m_rgpthrPendingSignalings = {0x0, 0x0, 0x0,
      0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0}, m_lePendingSignalingsOverflowList = {Flink = 0x80569f0, Blink = 0x80569f0}},
  suspensionInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e50948 <vtable for CorUnix::CThreadSuspensionInfo+8>}, m_fPending = 0x0,
---Type <return> to continue, or q <return> to quit---
    m_fSelfsusp = 0x0, m_fSuspendedForShutdown = 0x0, m_nBlockingPipe = 0xffffffff, m_lNumThreadsSuspendedByThisThread = {
      m_val = 0x0}, m_ptmSuspmutex = {__data = {__lock = 0x0, __count = 0x0, __owner = 0x0, __kind = 0x0, __nusers = 0x0, {
          __elision_data = {__espins = 0x0, __elision = 0x0}, __list = {__next = 0x0}}}, __size = {
        0x0 <repeats 24 times>}, __align = 0x0}, m_fSuspmutexInitialized = 0x1, m_semSusp = {__size = {
        0x0 <repeats 16 times>}, __align = 0x0}, m_semResume = {__size = {0x0 <repeats 16 times>}, __align = 0x0},
    m_fSemaphoresInitialized = 0x1}, sehInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e508e8 <vtable for CorUnix::CThreadSEHInfo+8>}, safe_state = 0x0,
    signal_code = 0x0}, tlsInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e50970 <vtable for CorUnix::CThreadTLSInfo+8>}, tlsSlots = {
      0x0 <repeats 64 times>}}, apcInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e50904 <vtable for CorUnix::CThreadApcInfo+8>}, m_ptainHead = 0x0,
    m_ptainTail = 0x0}, crtInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e50924 <vtable for CorUnix::CThreadCRTInfo+8>}, strtokContext = 0x0,
    wcstokContext = 0x0, localtimeBuffer = {tm_sec = 0x0, tm_min = 0x0, tm_hour = 0x0, tm_mday = 0x0, tm_mon = 0x0,
      tm_year = 0x0, tm_wday = 0x0, tm_yday = 0x0, tm_isdst = 0x0}, ctimeBuffer = {0x0 <repeats 26 times>}, ECVTBuffer = {
      0x0 <repeats 357 times>}}}
(gdb) c
Continuing.

Breakpoint 3, CatchHardwareExceptionHolder::~CatchHardwareExceptionHolder (this=0xbffff250)
    at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:370
370         CPalThread *pThread = InternalGetCurrentThread();
(gdb) n
371         --pThread->m_hardwareExceptionHolderCount;
(gdb) p pThread
$5 = (CorUnix::CPalThread *) 0x80566c8
(gdb) p/x *pThread
$6 = {_vptr$CPalThread = 0xb7e508d0 <vtable for CorUnix::CPalThread+8>, m_pNext = 0x0, m_dwExitCode = 0x103,
  m_fExitCodeSet = 0x0, m_csLock = {DebugInfo = 0x8056d40, LockCount = 0x0, RecursionCount = 0x0, OwningThread = 0x0,
    LockSemaphore = 0x0, SpinCount = 0x0, bInternal = 0x1, dwInitState = 0x1, csnds = {rgNativeDataStorage = {
        0x0 <repeats 76 times>}, pvAlign = 0x0}}, m_fLockInitialized = 0x1, m_fIsDummy = 0x0, m_lRefCount = 0x1,
  m_pThreadObject = 0x805a420, m_threadId = 0x6fe0, m_dwLwpId = 0x0, m_pthreadSelf = 0xb78a4740,
  m_hardwareExceptionHolderCount = 0x1, m_lpStartAddress = 0x0, m_lpStartParameter = 0x0, m_bCreateSuspended = 0x0,
  m_iThreadPriority = 0x0, m_eThreadType = 0x0, m_startMutex = {__data = {__lock = 0x0, __count = 0x0, __owner = 0x0,
      __kind = 0x0, __nusers = 0x0, {__elision_data = {__espins = 0x0, __elision = 0x0}, __list = {__next = 0x0}}},
    __size = {0x0 <repeats 24 times>}, __align = 0x0}, m_startCond = {__data = {__lock = 0x0, __futex = 0x0,
      __total_seq = 0x0, __wakeup_seq = 0x0, __woken_seq = 0x0, __mutex = 0x0, __nwaiters = 0x0, __broadcast_seq = 0x0},
    __size = {0x0 <repeats 48 times>}, __align = 0x0}, m_fStartItemsInitialized = 0x1, m_fStartStatus = 0x0,
  m_fStartStatusSet = 0x0, m_stackBase = 0x0, m_stackLimit = 0x0, static s_dwDefaultThreadStackSize = 0x40000,
  m_fInPal = 0x1, synchronizationInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e507dc <vtable for CorUnix::CThreadSynchronizationInfo+8>},
    m_tsThreadState = 0x1, m_shridWaitAwakened = 0x70, m_lLocalSynchLockCount = {m_val = 0x0}, m_lSharedSynchLockCount = {
      m_val = 0x0}, m_leOwnedObjsList = {Flink = 0x80567e0, Blink = 0x80567e0}, m_ownedNamedMutexListLock = {
      DebugInfo = 0x8056d20, LockCount = 0x0, RecursionCount = 0x0, OwningThread = 0x0, LockSemaphore = 0x0,
      SpinCount = 0x0, bInternal = 0x0, dwInitState = 0x1, csnds = {rgNativeDataStorage = {0x0 <repeats 76 times>},
        pvAlign = 0x0}}, m_ownedNamedMutexListHead = 0x0, m_tnwdNativeData = {mutex = {__data = {__lock = 0x0,
          __count = 0x0, __owner = 0x0, __kind = 0x0, __nusers = 0x0, {__elision_data = {__espins = 0x0, __elision = 0x0},
            __list = {__next = 0x0}}}, __size = {0x0 <repeats 24 times>}, __align = 0x0}, cond = {__data = {__lock = 0x0,
          __futex = 0x0, __total_seq = 0x0, __wakeup_seq = 0x0, __woken_seq = 0x0, __mutex = 0x0, __nwaiters = 0x0,
          __broadcast_seq = 0x0}, __size = {0x0 <repeats 48 times>}, __align = 0x0}, iPred = 0x0, dwObjectIndex = 0x0,
      twrWakeupReason = 0x0, fInitialized = 0x1}, m_twiWaitInfo = {wtWaitType = 0x0, wdWaitDomain = 0x0, lObjCount = 0x0,
      lSharedObjCount = 0x0, pthrOwner = 0x80566c8, rgpWTLNodes = {0x0 <repeats 64 times>}},
    static PendingSignalingsArraySize = 0xa, m_lPendingSignalingCount = 0x0, m_rgpthrPendingSignalings = {0x0, 0x0, 0x0,
      0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0}, m_lePendingSignalingsOverflowList = {Flink = 0x80569f0, Blink = 0x80569f0}},
  suspensionInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e50948 <vtable for CorUnix::CThreadSuspensionInfo+8>}, m_fPending = 0x0,
---Type <return> to continue, or q <return> to quit---
    m_fSelfsusp = 0x0, m_fSuspendedForShutdown = 0x0, m_nBlockingPipe = 0xffffffff, m_lNumThreadsSuspendedByThisThread = {
      m_val = 0x0}, m_ptmSuspmutex = {__data = {__lock = 0x0, __count = 0x0, __owner = 0x0, __kind = 0x0, __nusers = 0x0, {
          __elision_data = {__espins = 0x0, __elision = 0x0}, __list = {__next = 0x0}}}, __size = {
        0x0 <repeats 24 times>}, __align = 0x0}, m_fSuspmutexInitialized = 0x1, m_semSusp = {__size = {
        0x0 <repeats 16 times>}, __align = 0x0}, m_semResume = {__size = {0x0 <repeats 16 times>}, __align = 0x0},
    m_fSemaphoresInitialized = 0x1}, sehInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e508e8 <vtable for CorUnix::CThreadSEHInfo+8>}, safe_state = 0x0,
    signal_code = 0x0}, tlsInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e50970 <vtable for CorUnix::CThreadTLSInfo+8>}, tlsSlots = {
      0x0 <repeats 64 times>}}, apcInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e50904 <vtable for CorUnix::CThreadApcInfo+8>}, m_ptainHead = 0x0,
    m_ptainTail = 0x0}, crtInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e50924 <vtable for CorUnix::CThreadCRTInfo+8>}, strtokContext = 0x0,
    wcstokContext = 0x0, localtimeBuffer = {tm_sec = 0x0, tm_min = 0x0, tm_hour = 0x0, tm_mday = 0x0, tm_mon = 0x0,
      tm_year = 0x0, tm_wday = 0x0, tm_yday = 0x0, tm_isdst = 0x0}, ctimeBuffer = {0x0 <repeats 26 times>}, ECVTBuffer = {
      0x0 <repeats 357 times>}}}
(gdb) c
Continuing.
DLLTest2 PASSED
Starting pal_sxs test1 DllTest1

Breakpoint 2, CatchHardwareExceptionHolder::CatchHardwareExceptionHolder (this=0xbffff250)
    at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:364
364         CPalThread *pThread = InternalGetCurrentThread();
(gdb) n
365         ++pThread->m_hardwareExceptionHolderCount;
(gdb) p pThread
$7 = (CorUnix::CPalThread *) 0x80519e0
(gdb) p/x *pThread
$8 = {_vptr$CPalThread = 0xb7fc88d0 <vtable for CorUnix::CPalThread+8>, m_pNext = 0x0, m_dwExitCode = 0x103,
  m_fExitCodeSet = 0x0, m_csLock = {DebugInfo = 0x8052058, LockCount = 0x0, RecursionCount = 0x0, OwningThread = 0x0,
    LockSemaphore = 0x0, SpinCount = 0x0, bInternal = 0x1, dwInitState = 0x1, csnds = {rgNativeDataStorage = {
        0x0 <repeats 76 times>}, pvAlign = 0x0}}, m_fLockInitialized = 0x1, m_fIsDummy = 0x0, m_lRefCount = 0x1,
  m_pThreadObject = 0x8055750, m_threadId = 0x6fe0, m_dwLwpId = 0x0, m_pthreadSelf = 0xb78a4740,
  m_hardwareExceptionHolderCount = 0x0, m_lpStartAddress = 0x0, m_lpStartParameter = 0x0, m_bCreateSuspended = 0x0,
  m_iThreadPriority = 0x0, m_eThreadType = 0x0, m_startMutex = {__data = {__lock = 0x0, __count = 0x0, __owner = 0x0,
      __kind = 0x0, __nusers = 0x0, {__elision_data = {__espins = 0x0, __elision = 0x0}, __list = {__next = 0x0}}},
    __size = {0x0 <repeats 24 times>}, __align = 0x0}, m_startCond = {__data = {__lock = 0x0, __futex = 0x0,
      __total_seq = 0x0, __wakeup_seq = 0x0, __woken_seq = 0x0, __mutex = 0x0, __nwaiters = 0x0, __broadcast_seq = 0x0},
    __size = {0x0 <repeats 48 times>}, __align = 0x0}, m_fStartItemsInitialized = 0x1, m_fStartStatus = 0x0,
  m_fStartStatusSet = 0x0, m_stackBase = 0x0, m_stackLimit = 0x0, static s_dwDefaultThreadStackSize = 0x40000,
  m_fInPal = 0x1, synchronizationInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7fc87dc <vtable for CorUnix::CThreadSynchronizationInfo+8>},
    m_tsThreadState = 0x1, m_shridWaitAwakened = 0x70, m_lLocalSynchLockCount = {m_val = 0x0}, m_lSharedSynchLockCount = {
      m_val = 0x0}, m_leOwnedObjsList = {Flink = 0x8051af8, Blink = 0x8051af8}, m_ownedNamedMutexListLock = {
      DebugInfo = 0x8052038, LockCount = 0x0, RecursionCount = 0x0, OwningThread = 0x0, LockSemaphore = 0x0,
      SpinCount = 0x0, bInternal = 0x0, dwInitState = 0x1, csnds = {rgNativeDataStorage = {0x0 <repeats 76 times>},
        pvAlign = 0x0}}, m_ownedNamedMutexListHead = 0x0, m_tnwdNativeData = {mutex = {__data = {__lock = 0x0,
          __count = 0x0, __owner = 0x0, __kind = 0x0, __nusers = 0x0, {__elision_data = {__espins = 0x0, __elision = 0x0},
            __list = {__next = 0x0}}}, __size = {0x0 <repeats 24 times>}, __align = 0x0}, cond = {__data = {__lock = 0x0,
          __futex = 0x0, __total_seq = 0x0, __wakeup_seq = 0x0, __woken_seq = 0x0, __mutex = 0x0, __nwaiters = 0x0,
          __broadcast_seq = 0x0}, __size = {0x0 <repeats 48 times>}, __align = 0x0}, iPred = 0x0, dwObjectIndex = 0x0,
      twrWakeupReason = 0x0, fInitialized = 0x1}, m_twiWaitInfo = {wtWaitType = 0x0, wdWaitDomain = 0x0, lObjCount = 0x0,
      lSharedObjCount = 0x0, pthrOwner = 0x80519e0, rgpWTLNodes = {0x0 <repeats 64 times>}},
    static PendingSignalingsArraySize = 0xa, m_lPendingSignalingCount = 0x0, m_rgpthrPendingSignalings = {0x0, 0x0, 0x0,
      0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0}, m_lePendingSignalingsOverflowList = {Flink = 0x8051d08, Blink = 0x8051d08}},
  suspensionInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7fc8948 <vtable for CorUnix::CThreadSuspensionInfo+8>}, m_fPending = 0x0,
---Type <return> to continue, or q <return> to quit---
    m_fSelfsusp = 0x0, m_fSuspendedForShutdown = 0x0, m_nBlockingPipe = 0xffffffff, m_lNumThreadsSuspendedByThisThread = {
      m_val = 0x0}, m_ptmSuspmutex = {__data = {__lock = 0x0, __count = 0x0, __owner = 0x0, __kind = 0x0, __nusers = 0x0, {
          __elision_data = {__espins = 0x0, __elision = 0x0}, __list = {__next = 0x0}}}, __size = {
        0x0 <repeats 24 times>}, __align = 0x0}, m_fSuspmutexInitialized = 0x1, m_semSusp = {__size = {
        0x0 <repeats 16 times>}, __align = 0x0}, m_semResume = {__size = {0x0 <repeats 16 times>}, __align = 0x0},
    m_fSemaphoresInitialized = 0x1}, sehInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7fc88e8 <vtable for CorUnix::CThreadSEHInfo+8>}, safe_state = 0x0,
    signal_code = 0x0}, tlsInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7fc8970 <vtable for CorUnix::CThreadTLSInfo+8>}, tlsSlots = {
      0x0 <repeats 64 times>}}, apcInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7fc8904 <vtable for CorUnix::CThreadApcInfo+8>}, m_ptainHead = 0x0,
    m_ptainTail = 0x0}, crtInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7fc8924 <vtable for CorUnix::CThreadCRTInfo+8>}, strtokContext = 0x0,
    wcstokContext = 0x0, localtimeBuffer = {tm_sec = 0x0, tm_min = 0x0, tm_hour = 0x0, tm_mday = 0x0, tm_mon = 0x0,
      tm_year = 0x0, tm_wday = 0x0, tm_yday = 0x0, tm_isdst = 0x0}, ctimeBuffer = {0x0 <repeats 26 times>}, ECVTBuffer = {
      0x0 <repeats 357 times>}}}
(gdb) c
Continuing.

Program received signal SIGSEGV, Segmentation fault.
0xb7e641df in FailingFunction (p=0x11)
    at /home/maxwell/netcore/coreclr/src/pal/tests/palsuite/exception_handling/pal_sxs/test1/dlltest1.cpp:31
31          *p = 1;          // Causes an access violation exception
(gdb) c
Continuing.

Breakpoint 1, SEHProcessException (exception=0xbfffebb0) at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:292
292             if (CatchHardwareExceptionHolder::IsEnabled())
(gdb) s
CatchHardwareExceptionHolder::IsEnabled () at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:376
376         CPalThread *pThread = InternalGetCurrentThread();
(gdb) n
377         return pThread->IsHardwareExceptionsEnabled();
(gdb) p pThread
$9 = (CorUnix::CPalThread *) 0x80566c8
(gdb) p/x *pThread
$10 = {_vptr$CPalThread = 0xb7e508d0 <vtable for CorUnix::CPalThread+8>, m_pNext = 0x0, m_dwExitCode = 0x103,
  m_fExitCodeSet = 0x0, m_csLock = {DebugInfo = 0x8056d40, LockCount = 0x0, RecursionCount = 0x0, OwningThread = 0x0,
    LockSemaphore = 0x0, SpinCount = 0x0, bInternal = 0x1, dwInitState = 0x1, csnds = {rgNativeDataStorage = {
        0x0 <repeats 76 times>}, pvAlign = 0x0}}, m_fLockInitialized = 0x1, m_fIsDummy = 0x0, m_lRefCount = 0x1,
  m_pThreadObject = 0x805a420, m_threadId = 0x6fe0, m_dwLwpId = 0x0, m_pthreadSelf = 0xb78a4740,
  m_hardwareExceptionHolderCount = 0x0, m_lpStartAddress = 0x0, m_lpStartParameter = 0x0, m_bCreateSuspended = 0x0,
  m_iThreadPriority = 0x0, m_eThreadType = 0x0, m_startMutex = {__data = {__lock = 0x0, __count = 0x0, __owner = 0x0,
      __kind = 0x0, __nusers = 0x0, {__elision_data = {__espins = 0x0, __elision = 0x0}, __list = {__next = 0x0}}},
    __size = {0x0 <repeats 24 times>}, __align = 0x0}, m_startCond = {__data = {__lock = 0x0, __futex = 0x0,
      __total_seq = 0x0, __wakeup_seq = 0x0, __woken_seq = 0x0, __mutex = 0x0, __nwaiters = 0x0, __broadcast_seq = 0x0},
    __size = {0x0 <repeats 48 times>}, __align = 0x0}, m_fStartItemsInitialized = 0x1, m_fStartStatus = 0x0,
  m_fStartStatusSet = 0x0, m_stackBase = 0x0, m_stackLimit = 0x0, static s_dwDefaultThreadStackSize = 0x40000,
  m_fInPal = 0x1, synchronizationInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e507dc <vtable for CorUnix::CThreadSynchronizationInfo+8>},
    m_tsThreadState = 0x1, m_shridWaitAwakened = 0x70, m_lLocalSynchLockCount = {m_val = 0x0}, m_lSharedSynchLockCount = {
      m_val = 0x0}, m_leOwnedObjsList = {Flink = 0x80567e0, Blink = 0x80567e0}, m_ownedNamedMutexListLock = {
      DebugInfo = 0x8056d20, LockCount = 0x0, RecursionCount = 0x0, OwningThread = 0x0, LockSemaphore = 0x0,
      SpinCount = 0x0, bInternal = 0x0, dwInitState = 0x1, csnds = {rgNativeDataStorage = {0x0 <repeats 76 times>},
        pvAlign = 0x0}}, m_ownedNamedMutexListHead = 0x0, m_tnwdNativeData = {mutex = {__data = {__lock = 0x0,
          __count = 0x0, __owner = 0x0, __kind = 0x0, __nusers = 0x0, {__elision_data = {__espins = 0x0, __elision = 0x0},
            __list = {__next = 0x0}}}, __size = {0x0 <repeats 24 times>}, __align = 0x0}, cond = {__data = {__lock = 0x0,
          __futex = 0x0, __total_seq = 0x0, __wakeup_seq = 0x0, __woken_seq = 0x0, __mutex = 0x0, __nwaiters = 0x0,
          __broadcast_seq = 0x0}, __size = {0x0 <repeats 48 times>}, __align = 0x0}, iPred = 0x0, dwObjectIndex = 0x0,
      twrWakeupReason = 0x0, fInitialized = 0x1}, m_twiWaitInfo = {wtWaitType = 0x0, wdWaitDomain = 0x0, lObjCount = 0x0,
      lSharedObjCount = 0x0, pthrOwner = 0x80566c8, rgpWTLNodes = {0x0 <repeats 64 times>}},
    static PendingSignalingsArraySize = 0xa, m_lPendingSignalingCount = 0x0, m_rgpthrPendingSignalings = {0x0, 0x0, 0x0,
      0x0, 0x0, 0x0, 0x0, 0x0, 0x0, 0x0}, m_lePendingSignalingsOverflowList = {Flink = 0x80569f0, Blink = 0x80569f0}},
  suspensionInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e50948 <vtable for CorUnix::CThreadSuspensionInfo+8>}, m_fPending = 0x0,
---Type <return> to continue, or q <return> to quit---
    m_fSelfsusp = 0x0, m_fSuspendedForShutdown = 0x0, m_nBlockingPipe = 0xffffffff, m_lNumThreadsSuspendedByThisThread = {
      m_val = 0x0}, m_ptmSuspmutex = {__data = {__lock = 0x0, __count = 0x0, __owner = 0x0, __kind = 0x0, __nusers = 0x0, {
          __elision_data = {__espins = 0x0, __elision = 0x0}, __list = {__next = 0x0}}}, __size = {
        0x0 <repeats 24 times>}, __align = 0x0}, m_fSuspmutexInitialized = 0x1, m_semSusp = {__size = {
        0x0 <repeats 16 times>}, __align = 0x0}, m_semResume = {__size = {0x0 <repeats 16 times>}, __align = 0x0},
    m_fSemaphoresInitialized = 0x1}, sehInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e508e8 <vtable for CorUnix::CThreadSEHInfo+8>}, safe_state = 0x0,
    signal_code = 0x0}, tlsInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e50970 <vtable for CorUnix::CThreadTLSInfo+8>}, tlsSlots = {
      0x0 <repeats 64 times>}}, apcInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e50904 <vtable for CorUnix::CThreadApcInfo+8>}, m_ptainHead = 0x0,
    m_ptainTail = 0x0}, crtInfo = {<CorUnix::CThreadInfoInitializer> = {
      _vptr$CThreadInfoInitializer = 0xb7e50924 <vtable for CorUnix::CThreadCRTInfo+8>}, strtokContext = 0x0,
    wcstokContext = 0x0, localtimeBuffer = {tm_sec = 0x0, tm_min = 0x0, tm_hour = 0x0, tm_mday = 0x0, tm_mon = 0x0,
      tm_year = 0x0, tm_wday = 0x0, tm_yday = 0x0, tm_isdst = 0x0}, ctimeBuffer = {0x0 <repeats 26 times>}, ECVTBuffer = {
      0x0 <repeats 357 times>}}}
(gdb) c
Continuing.

Program received signal SIGSEGV, Segmentation fault.
0xbfffecb8 in ?? ()

[janvorli]

@seanshpark I am sorry, I can see I was not clear. I didn't mean dumping the members of the pThread, but getting the current Linux thread id using the gdb. You can e.g. run "info threads" - the current one is marked in the list.

[seanshpark]

not a problem :)

(gdb) r
Starting program: /home/maxwell/netcore/coreclr/bin/obj/Linux.x86.Debug/src/pal/tests/palsuite/exception_handling/pal_sxs/test1/paltest_pal_sxs_test1
[Thread debugging using libthread_db enabled]
Using host libthread_db library "/lib/i386-linux-gnu/libthread_db.so.1".
PAL_SXS test1 SIGSEGV handler (nil)

Breakpoint 5, CorUnix::CreateThreadObject (pThread=0x80519e0, pNewThread=0x80519e0, phThread=0xbfffee84)
    at /home/maxwell/netcore/coreclr/src/pal/src/thread/thread.cpp:1968
1968        *phThread = hThread;
(gdb) c
Continuing.

Breakpoint 5, CorUnix::CreateThreadObject (pThread=0x80566c8, pNewThread=0x80566c8, phThread=0xbfffee84)
    at /home/maxwell/netcore/coreclr/src/pal/src/thread/thread.cpp:1968
1968        *phThread = hThread;
(gdb) c
Continuing.
Starting pal_sxs test1 DllTest2

Breakpoint 2, CatchHardwareExceptionHolder::CatchHardwareExceptionHolder (this=0xbffff250)
    at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:364
364         CPalThread *pThread = InternalGetCurrentThread();
(gdb) n
365         ++pThread->m_hardwareExceptionHolderCount;
(gdb) info threads
  Id   Target Id         Frame
* 1    Thread 0xb78a4740 (LWP 28758) "paltest_pal_sxs" CatchHardwareExceptionHolder::CatchHardwareExceptionHolder (
    this=0xbffff250) at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:365
(gdb) c
Continuing.

Program received signal SIGSEGV, Segmentation fault.
0xb7cec1df in FailingFunction (p=0x22)
    at /home/maxwell/netcore/coreclr/src/pal/tests/palsuite/exception_handling/pal_sxs/test1/dlltest2.cpp:31
31          *p = 1;          // Causes an access violation exception
(gdb) c
Continuing.

Breakpoint 1, SEHProcessException (exception=0xbfffebb0) at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:292
292             if (CatchHardwareExceptionHolder::IsEnabled())
(gdb) s
CatchHardwareExceptionHolder::IsEnabled () at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:376
376         CPalThread *pThread = InternalGetCurrentThread();
(gdb) n
377         return pThread->IsHardwareExceptionsEnabled();
(gdb) info threads
  Id   Target Id         Frame
* 1    Thread 0xb78a4740 (LWP 28758) "paltest_pal_sxs" CatchHardwareExceptionHolder::IsEnabled ()
    at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:377
(gdb) c
Continuing.

Breakpoint 3, CatchHardwareExceptionHolder::~CatchHardwareExceptionHolder (this=0xbffff250)
    at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:370
370         CPalThread *pThread = InternalGetCurrentThread();
(gdb) c
Continuing.
DLLTest2 PASSED
Starting pal_sxs test1 DllTest1

Breakpoint 2, CatchHardwareExceptionHolder::CatchHardwareExceptionHolder (this=0xbffff250)
    at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:364
364         CPalThread *pThread = InternalGetCurrentThread();
(gdb) c
Continuing.

Program received signal SIGSEGV, Segmentation fault.
0xb7e641df in FailingFunction (p=0x11)
    at /home/maxwell/netcore/coreclr/src/pal/tests/palsuite/exception_handling/pal_sxs/test1/dlltest1.cpp:31
31          *p = 1;          // Causes an access violation exception
(gdb) c
Continuing.

Breakpoint 1, SEHProcessException (exception=0xbfffebb0) at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:292
292             if (CatchHardwareExceptionHolder::IsEnabled())
(gdb) s
CatchHardwareExceptionHolder::IsEnabled () at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:376
376         CPalThread *pThread = InternalGetCurrentThread();
(gdb) n
377         return pThread->IsHardwareExceptionsEnabled();
(gdb) info threads
  Id   Target Id         Frame
* 1    Thread 0xb78a4740 (LWP 28758) "paltest_pal_sxs" CatchHardwareExceptionHolder::IsEnabled ()
    at /home/maxwell/netcore/coreclr/src/pal/src/exception/seh.cpp:377
(gdb) c
Continuing.

Program received signal SIGSEGV, Segmentation fault.
0xbfffecb8 in ?? ()