Emacs from HEAD hangs on M$ (built with MinGW)

Emacs from HEAD hangs on M$ (built with MinGW)

[dhruva]

Hi,
 I just did a 'make bootstrap', emacs freezes but consumes 45% CPU and becomes non-responsive. I have build using GCC (MinGW) and running it on WXP.

Here is the back trace:
Thread 3 (thread 3692.0x1234):
#0  0x7c810659 in KERNEL32!CreateThread () from C:\WINDOWS\system32\kernel32.dll

Thread 2 (thread 3692.0x90c):
#0  0x7c90eb94 in ntdll!LdrAccessResource () from C:\WINDOWS\system32\ntdll.dll
#1  0x7e4191be in USER32!GetProcessWindowStation () from C:\WINDOWS\system32\user32.dll
#2  0x7e42e042 in USER32!GetMessageA () from C:\WINDOWS\system32\user32.dll
#3  0x011c71ee in w32_msg_pump (msg_buf=0x20dbff5c) at w32fns.c:2481
#4  0x011c73ad in w32_msg_worker (arg=0x0) at w32fns.c:2700
#5  0x7c80b683 in KERNEL32!GetModuleFileNameA () from C:\WINDOWS\system32\kernel32.dll
#6  0x00000000 in ?? ()

Thread 1 (thread 3692.0x15d4):
#0  0x7c90eb94 in ntdll!LdrAccessResource () from C:\WINDOWS\system32\ntdll.dll
#1  0x7c90e9ab in ntdll!ZwWaitForMultipleObjects () from C:\WINDOWS\system32\ntdll.dll
#2  0x7c8094e2 in KERNEL32!CreateFileMappingA () from C:\WINDOWS\system32\kernel32.dll
#3  0x7e4195f9 in USER32!GetLastInputInfo () from C:\WINDOWS\system32\user32.dll
#4  0x7e4196a8 in USER32!MsgWaitForMultipleObjects () from C:\WINDOWS\system32\user32.dll
#5  0x01089571 in sys_select (nfds=1, rfds=0x82f5a8, wfds=0x0, efds=0x0, timeout=0x82f5a0) at w32proc.c:1270
#6  0x01083703 in wait_reading_process_output (time_limit=0, microsecs=0, read_kbd=-1, do_display=1, wait_for_cell=44644353, wait_proc=0x0, just_wait_proc=0) at process.c:4806
#7  0x0100b58a in kbd_buffer_get_event (kbp=0x82f720, used_mouse_menu=0x82fa44, end_time=0x0) at keyboard.c:4041
#8  0x01009878 in read_char (commandflag=1, nmaps=3, maps=0x82f8a0, prev_event=44644353, used_mouse_menu=0x82fa44, end_time=0x0) at keyboard.c:3004
#9  0x01012c2a in read_key_sequence (keybuf=0x82fc58, bufsize=30, prompt=44644353, dont_downcase_last=0, can_return_switch_frame=1, fix_current_buffer=1) at keyboard.c:9349
#10 0x010063fa in command_loop_1 () at keyboard.c:1620
#11 0x01021a56 in internal_condition_case (bfun=0x10060cf <command_loop_1>, handlers=44708081, hfun=0x1005acc <cmd_error>) at eval.c:1511
#12 0x01005e18 in command_loop_2 () at keyboard.c:1338
#13 0x01021566 in internal_catch (tag=44704153, func=0x1005df8 <command_loop_2>, arg=44644353) at eval.c:1247
#14 0x01005dcf in command_loop () at keyboard.c:1317
#15 0x010056d2 in recursive_edit_1 () at keyboard.c:942
#16 0x01005840 in Frecursive_edit () at keyboard.c:1004
#17 0x010027dc in main (argc=2, argv=0xa343a8) at emacs.c:1687
#0  0x7c810659 in KERNEL32!CreateThread () from C:\WINDOWS\system32\kernel32.dll

I continue, emacs window is white washed, I type C-g to get back initial screen and interrupt again to take a trace:
Thread 4 (thread 3692.0x16e8):
#0  0x7c810659 in KERNEL32!CreateThread () from C:\WINDOWS\system32\kernel32.dll
Cannot access memory at address 0x3

Thread 2 (thread 3692.0x90c):
#0  0x7c90eb94 in ntdll!LdrAccessResource () from C:\WINDOWS\system32\ntdll.dll
#1  0x7e4191be in USER32!GetProcessWindowStation () from C:\WINDOWS\system32\user32.dll
#2  0x7e42e042 in USER32!GetMessageA () from C:\WINDOWS\system32\user32.dll
#3  0x011c71ee in w32_msg_pump (msg_buf=0x20dbff5c) at w32fns.c:2481
#4  0x011c73ad in w32_msg_worker (arg=0x0) at w32fns.c:2700
#5  0x7c80b683 in KERNEL32!GetModuleFileNameA () from C:\WINDOWS\system32\kernel32.dll
#6  0x00000000 in ?? ()

Thread 1 (thread 3692.0x15d4):
#0  wait_reading_process_output (time_limit=0, microsecs=0, read_kbd=-1, do_display=1, wait_for_cell=44644353, wait_proc=0x0, just_wait_proc=0) at process.c:4830
#1  0x0100b58a in kbd_buffer_get_event (kbp=0x82f720, used_mouse_menu=0x82fa44, end_time=0x0) at keyboard.c:4041
#2  0x01009878 in read_char (commandflag=1, nmaps=3, maps=0x82f8a0, prev_event=44644353, used_mouse_menu=0x82fa44, end_time=0x0) at keyboard.c:3004
#3  0x01012c2a in read_key_sequence (keybuf=0x82fc58, bufsize=30, prompt=44644353, dont_downcase_last=0, can_return_switch_frame=1, fix_current_buffer=1) at keyboard.c:9349
#4  0x010063fa in command_loop_1 () at keyboard.c:1620
#5  0x01021a56 in internal_condition_case (bfun=0x10060cf <command_loop_1>, handlers=44708081, hfun=0x1005acc <cmd_error>) at eval.c:1511
#6  0x01005e18 in command_loop_2 () at keyboard.c:1338
#7  0x01021566 in internal_catch (tag=44704153, func=0x1005df8 <command_loop_2>, arg=44644353) at eval.c:1247
#8  0x01005dcf in command_loop () at keyboard.c:1317
#9  0x010056d2 in recursive_edit_1 () at keyboard.c:942
#10 0x01005840 in Frecursive_edit () at keyboard.c:1004
#11 0x010027dc in main (argc=2, argv=0xa343a8) at emacs.c:1687
#0  0x7c810659 in KERNEL32!CreateThread () from C:\WINDOWS\system32\kernel32.dll

I feel the keyboard is not being read except for C-g sequence. I was building emacs with MSVC 2003 till yesterday and had to upgrade to MSVC 8 with which emacs does not build. I therefore decided to use GCC (MinGW) to build and started seeing these issues today (may not be related to GCC). I can switch back to MSVC 2003 as I have removed it!

-dhruva



      Unlimited freedom, unlimited storage. Get it now, on http://help.yahoo.com/l/in/yahoo/mail/yahoomail/tools/tools-08.html/

Re: Emacs from HEAD hangs on M$ (built with MinGW)

[Lennart Borgman (gmail)]

dhruva wrote:
> Hi,
>  I just did a 'make bootstrap', emacs freezes but consumes 45% CPU
> and becomes non-responsive. I have build using GCC (MinGW) and
> running it on WXP.

I see the same thing.

> I feel the keyboard is not being read except for C-g sequence.
 > I was building emacs with MSVC 2003 till yesterday and had to
 > upgrade to MSVC 8 with which emacs does not build. I therefore
 > decided to use GCC (MinGW) to build and started seeing these
> issues today (may not be related to GCC). I can switch back to
> MSVC 2003 as I have removed it!

I think this is not related to GCC. Jason did some changes to finally 
get the display update on w32 to work fast. I believe he is using MingGW.

Jason did his checkin just some hours before Dan N did his.

It looks like it could be Dan's changes that has broken something on w32.

Re: Emacs from HEAD hangs on M$ (built with MinGW)

[Jason Rumney]

Lennart Borgman (gmail) wrote:

> I think this is not related to GCC. Jason did some changes to finally
> get the display update on w32 to work fast. I believe he is using MingGW.
> 
> Jason did his checkin just some hours before Dan N did his.
> 
> It looks like it could be Dan's changes that has broken something on w32.

The following change looks suspicious:

http://cvs.savannah.gnu.org/viewvc/emacs/src/term.c?root=emacs&r1=1.222&r2=1.223

Re: Emacs from HEAD hangs on M$ (built with MinGW)

[Lennart Borgman (gmail)]

Jason Rumney wrote:
> Lennart Borgman (gmail) wrote:
> 
>> I think this is not related to GCC. Jason did some changes to finally
>> get the display update on w32 to work fast. I believe he is using MingGW.
>>
>> Jason did his checkin just some hours before Dan N did his.
>>
>> It looks like it could be Dan's changes that has broken something on w32.
> 
> The following change looks suspicious:
> 
> http://cvs.savannah.gnu.org/viewvc/emacs/src/term.c?root=emacs&r1=1.222&r2=1.223

Exactly what do you think is wrong there. It looks like Dan removed the 
test for MULTI_KBOARD and then just simplified the test for WINDOWSNT.

Re: Emacs from HEAD hangs on M$ (built with MinGW)

[Juanma Barranquero]

On Thu, Jul 31, 2008 at 17:47, Jason Rumney <jasonr@gnu.org> wrote:

> The following change looks suspicious:
>
> http://cvs.savannah.gnu.org/viewvc/emacs/src/term.c?root=emacs&r1=1.222&r2=1.223

It's in fact a typo in this one:

http://cvs.savannah.gnu.org/viewvc/emacs/emacs/src/keyboard.c?r1=1.967&r2=1.968

It is fixed now.

  Juanma

Re: Emacs from HEAD hangs on M$ (built with MinGW)

[Chong Yidong]

"Juanma Barranquero" <lekktu@gmail.com> writes:

> On Thu, Jul 31, 2008 at 17:47, Jason Rumney <jasonr@gnu.org> wrote:
>
>> The following change looks suspicious:
>>
>> http://cvs.savannah.gnu.org/viewvc/emacs/src/term.c?root=emacs&r1=1.222&r2=1.223
>
> It's in fact a typo in this one:
>
> http://cvs.savannah.gnu.org/viewvc/emacs/emacs/src/keyboard.c?r1=1.967&r2=1.968
>
> It is fixed now.

Thanks for finding it.