I cannot seem to build perl 5.6.2 in Cygwin 1.7.11.
Here are the steps I am following to do it so far.
Download perl 5.6.2 source from website and untar to some folder "perl-5.6.2"
cd perl-5.6.2/bld
sh ../Configure -des -Dmksymlinks -Dprefix=/common/ndd/perl/5.6.2 -DDEBUGGING 2>&1
make 2>&1
make install 2>&1
Here are the logs from this:
http://pastebin.com/pqLg4S7z
What happens is make creates "perl-5.6.2/bld/perl.exe" and this file generates an abort signal every time.
Here is the backtrace from running gdb on it. (Note: I configured with -DDEBUGGING so all the gcc debug flags should be enabled.)
(gdb) run
Starting program: /openlogic/build/work/perl-5.6.2/bld/perl.exe
[New Thread 3348.0xf54]
[New Thread 3348.0x48c]
Program received signal SIGABRT, Aborted.
0x00000000 in ?? ()
(gdb) backtrace
#0 0x00000000 in ?? ()
#1 0x7792f8b1 in ntdll!RtlUpdateClonedSRWLock ()
from /cygdrive/c/Windows/system32/ntdll.dll
#2 0x757f0a91 in WaitForSingleObjectEx ()
from /cygdrive/c/Windows/syswow64/KERNELBASE.dll
#3 0x000000a8 in ?? ()
#4 0x00000000 in ?? ()
(gdb) quit
Here id the ldd.exe command ran on perl.exe
$ ldd /ndipiazza/build/work/perl-5.6.2/bld/perl.exe
ntdll.dll => /cygdrive/c/Windows/SysWOW64/ntdll.dll (0x77910000)
kernel32.dll => /cygdrive/c/Windows/syswow64/kernel32.dll (0x75cb0000)
KERNELBASE.dll => /cygdrive/c/Windows/syswow64/KERNELBASE.dll (0x757e0000)
libperl5_6_2.dll => /ndipiazza/build/work/perl-5.6.2/bld/libperl5_6_2.dll (0x66140000)
cygcrypt-0.dll => /usr/bin/cygcrypt-0.dll (0x67db0000)
cygwin1.dll => /usr/bin/cygwin1.dll (0x61000000)
??? => ??? (0x570000)
Can anyone see why I would be getting this sigabrt?
EDIT: here is a link to a discussion going on in Cygwin mailing list: http://cygwin.com/ml/cygwin/2012-07/msg00368.html
5.6 is outdated for your build environment. Maintenance patches exist.
I recommend to install perlbrew and patchperl:
$ perlbrew install-patchperl
which will DTRT.
This question is above my pay grade and this is a shot in the dark, but the two things I remember that ruin a perl5.6.2 build on Cygwin are
spaces in the $PATH (e.g. /cygdrive/c/Program Files/BlahBlahBlah)
configure script unable to determine the signal names
Check your config.sh script and see whether it contains lines like
sig_count='1'
sig_name='ZERO '
sig_name_init='"ZERO", 0'
sig_num='0 '
sig_num_init='0, 0'
You want it to say something like
sig_count='33'
sig_name='ZERO HUP INT QUIT ILL TRAP ABRT EMT FPE KILL BUS SEGV SYS PIPE ALRM TERM URG STOP TSTP CONT CHLD TTIN TTOU IO XCPU XFSZ VTALRM PROF WINCH LOST USR1 USR2 RTMAX CLD POLL PWR RTMIN '
sig_name_init='"ZERO", "HUP", "INT", "QUIT", "ILL", "TRAP", "ABRT", "EMT", "FPE", "KILL", "BUS", "SEGV", "SYS", "PIPE", "ALRM", "TERM", "URG", "STOP", "TSTP", "CONT", "CHLD", "TTIN", "TTOU", "IO", "XCPU", "XFSZ", "VTALRM", "PROF", "WINCH", "LOST", "USR1", "USR2", "RTMAX", "CLD", "POLL", "PWR", "RTMIN", 0'
sig_num='0 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 20 23 29 32 '
sig_num_init='0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25, 26, 27, 28, 29, 30, 31, 32, 20, 23, 29, 32, 0'
(I copied these values from a config.sh from a build for a newer version of Perl. The values you need may vary slightly)
I figured it out. Perl 5.6.2 tries to use it's own malloc. This is incompatible with Cygwin's malloc.
I was given some help on Cygwin mailing list to fix this: http://cygwin.com/ml/cygwin/2012-07/msg00380.html
Bottom line, add -Dusemymalloc=n to the sh ./Configure arguments.
Then it works.
Related
I am using below command to start the training of deepspeech model
%cd /content/DeepSpeech
!python3 DeepSpeech.py \
--drop_source_layers 2 --scorer /content/DeepSpeech/data/lm/kenlm-nigerian.scorer\
--train_cudnn True --early_stop True --es_epochs 6 --n_hidden 2048 --epochs 5 \
--export_dir /content/models/ --checkpoint_dir /content/model_checkpoints/ \
--train_files /content/train.csv --dev_files /content/dev.csv --test_files /content/test.csv \
--learning_rate 0.0001 --train_batch_size 64 --test_batch_size 32 --dev_batch_size 32 --export_file_name 'he_model_5' \
--max_to_keep 3
I keep getting following error again and again.
(0) Invalid argument: Not enough time for target transition sequence (required: 28, available: 24)0You can turn this error into a warning by using the flag ignore_longer_outputs_than_inputs
(1) Invalid argument: Not enough time for target transition sequence (required: 28, available: 24)0You can turn this error into a warning by using the flag ignore_longer_outputs_than_inputs
Following worked for me
Go to
DeepSpeech/training/deepspeech_training/train.py
Now look for following particular line (Normally in 240-250)
total_loss = tfv1.nn.ctc_loss(labels=batch_y, inputs=logits, sequence_length=batch_seq_len)
Change it to as following
total_loss = tfv1.nn.ctc_loss(labels=batch_y, inputs=logits, sequence_length=batch_seq_len, )
I installed on CentOs successfully ever. However, here is another CentOs I used, and it failed to stared rabbitMq.
My erlang from here.
[rabbitmq-erlang]
name=rabbitmq-erlang
baseurl=https://dl.bintray.com/rabbitmq/rpm/erlang/20/el/7
gpgcheck=1
gpgkey=https://dl.bintray.com/rabbitmq/Keys/rabbitmq-release-signing-key.asc
repo_gpgcheck=0
enabled=1
this is my erl_crash.dump.
erl_crash_dump:0.5
Sat Jun 23 09:17:30 2018
Slogan: init terminating in do_boot ({error,{no such file or directory,asn1.app}})
System version: Erlang/OTP 20 [erts-9.3.3] [source] [64-bit] [smp:24:24] [ds:24:24:10] [async-threads:384] [hipe] [kernel-poll:true]
Compiled: Tue Jun 19 22:25:03 2018
Taints: erl_tracer,zlib
Atoms: 14794
Calling Thread: scheduler:2
=scheduler:1
Scheduler Sleep Info Flags: SLEEPING | TSE_SLEEPING | WAITING
Scheduler Sleep Info Aux Work:
Current Port:
Run Queue Max Length: 0
Run Queue High Length: 0
Run Queue Normal Length: 0
Run Queue Low Length: 0
Run Queue Port Length: 0
Run Queue Flags: OUT_OF_WORK | HALFTIME_OUT_OF_WORK
Current Process:
=scheduler:2
Scheduler Sleep Info Flags:
Scheduler Sleep Info Aux Work: THR_PRGR_LATER_OP
Current Port:
Run Queue Max Length: 0
Run Queue High Length: 0
Run Queue Normal Length: 0
Run Queue Low Length: 0
Run Queue Port Length: 0
Run Queue Flags: OUT_OF_WORK | HALFTIME_OUT_OF_WORK | NONEMPTY | EXEC
Current Process: <0.0.0>
Current Process State: Running
Current Process Internal State: ACT_PRIO_NORMAL | USR_PRIO_NORMAL | PRQ_PRIO_NORMAL | ACTIVE | RUNNING | TRAP_EXIT | ON_HEAP_MSGQ
Current Process Program counter: 0x00007fbd81fa59c0 (init:boot_loop/2 + 64)
Current Process CP: 0x0000000000000000 (invalid)
how to identify this problem ? Thank you.
A library that I'm using XML::Compile::Translate::Reader calls Log::Report's error method
or error __x"data for element or block starting with `{tag}' missing at {path}"
, tag => $label, path => $path, _class => 'misfit';
As I've got Log::Report set to debug mode, it returns a stack trace for an error.
[11 07 2014 22:17:39] [2804] error: data for element or block starting with `MSISDN' missing at {http://www.sigvalue.com/acc}TA
at /usr/local/share/perl5/XML/Compile/Translate/Reader.pm line 476
Log::Report::error("Log::Report::Message=HASH(0x2871cf8)") at /usr/local/share/perl5/XML/Compile/Translate/Reader.pm line 476
<snip>
XML::Compile::SOAP::Daemon::LWPutil::lwp_run_request("HTTP::Request=HASH(0x2882858)", "CODE(0x231ba38)", "HTTP::Daemon::ClientConn::SSL=GLOB(0x231b9c0)", undef) at /usr/local/share/perl5/XML/Compile/SOAP/Daemon/LWPutil.pm line 95
Any::Daemon::run("XML::Compile::SOAP::Daemon::AnyDaemon=HASH(0x7a3168)", "child_task", "CODE(0x2548128)", "max_childs", 36, "background", 1) at /usr/local/share/perl5/XML/Compile/SOAP/Daemon/AnyDaemon.pm line 75
XML::Compile::SOAP::Daemon::AnyDaemon::_run("XML::Compile::SOAP::Daemon::AnyDaemon=HASH(0x7a3168)", "HASH(0x18dda00)") at /usr/local/share/perl5/XML/Compile/SOAP/Daemon.pm line 99
(eval)("XML::Compile::SOAP::Daemon::AnyDaemon=HASH(0x7a3168)", "HASH(0x18dda00)") at /usr/local/share/perl5/XML/Compile/SOAP/Daemon.pm line 94
XML::Compile::SOAP::Daemon::run("XML::Compile::SOAP::Daemon::AnyDaemon=HASH(0x7a3168)", "name", "rizserver.pl", "background", 1, "max_childs", 36, "socket", [7 more]) at ./rizserver.pl line 95
There is lots of juicy data in those HASH, SCALAR, GLOB, and other elements that I want to get logged; as we are having trouble logging the original request in case it doesn't match.
I've explored using
Some leads that I don't know how to use are using Log::Dispatch, or some sort of Filter on Log::Report; but in the end, all I really want is to apply Data::Dumper to those elements.
I'm trying to print the data received on a socket - the contents of ubuf on the return of sys_recv. I cant get the %M format specifier to work properly. Can someone please explain how to use it properly. Thanks
stap -L 'kernel.function("sys_recv#net/socket.c")'
kernel.function("sys_recv#net/socket.c:1800") $fd:int $ubuf:void* $size:size_t $flags:unsigned int
using this probe:
[laris#kakitis stap]$ cat socket-recv.stp
#! /usr/bin/env stap
probe kernel.function("sys_recv#net/socket.c").return {
if (pid() == target())
printf ("%s fd %d size %d ubuf %p %10M \n ", ppfunc(),$fd,$size,$ubuf,$ubuf)
}
From my reading of the man page the format %10M should return 10 bytes from the location pointed to by $ubuf:void but I only get 1. Adjusting the parameter 10 shifts the one character output rather than showing more or less memory
[root#kakitis stap]# stap -x 16796 socket-recv.stp
sys_recv fd 13 size 64071 ubuf 0x86ceca0 70
sys_recv fd 13 size 62679 ubuf 0x86cf210 50
Changing 10 to 2 gives this
[root#kakitis stap]# stap -x 16796 socket-recv.stp
sys_recv fd 13 size 64071 ubuf 0x86ceca0 70
sys_recv fd 13 size 62679 ubuf 0x86cf210 50
System particulars are:
[laris#kakitis stap]$ stap --version
Systemtap translator/driver (version 2.1/0.154, rpm 2.1-2.fc17)
Copyright (C) 2005-2013 Red Hat, Inc. and others
This is free software; see the source for copying conditions.
enabled features: AVAHI LIBRPM LIBSQLITE3 NSS TR1_UNORDERED_MAP NLS
[laris#kakitis stap]$ uname -a
Linux kakitis 3.4.33 #1 SMP Tue Jan 7 14:15:58 EST 2014 i686 i686 i386 GNU/Linux
[laris#kakitis stap]$ cat /etc/redhat-release
Fedora release 17 (Beefy Miracle)
Don't confuse the output-width and precision parameters for printf(). The following will do what you meant:
printf ("%33.10M", $pointer)
to print 10 bytes (20 hex characters) in a 33-character-wide output field. One or both numbers can be replaced by *, so that the respective widths are passed as parameters before the $pointer. The upstream man page has been updated with an example.
My process is hung on solaris.
following is the the output of truss command.
bash-3.00# truss -p 14350
write(1, 0x10057BD24, 4) (sleeping...)
Any ideas on why write is going into sleep.
following is the kernel stack.
bash-3.00# mdb -k
Loading modules: [ unix genunix specfs dtrace ufs pcisch ip hook neti sctp arp usba s1394 fcp fctl nca lofs zfs md audiosup sd cpc random crypto wrsmd fcip logindmux ptm sppp nfs ipc ]
> 0t14350::pid2proc |::walk thread |::findstack -v
stack pointer for thread 3000230c1c0: 2a102192fc1
[ 000002a102192fc1 cv_wait_sig_swap_core+0x130() ]
000002a102193071 fifo_write+0xdc(4000, 2a102193a98, 20a1, 0, 0, 600154582e0)
000002a102193131 fop_write+0x20(60015370780, 2a102193a98, 0, 3000283c910, 0, 127a574)
000002a1021931e1 write+0x268(1, 8058, 60015552268, 4, 3, 1)
000002a1021932e1 syscall_trap+0xac(1, 10057bd24, 4, 10057bd24, 10057d124, 10a)
>