Re: Last version of of citrix Client core dumps (Was Re: How to run Microsoft Internet Explorer on NetBSD?)

[Patrick Welche]

On Fri, Feb 26, 2016 at 04:11:53PM +, Christos Zoulas wrote:
> In article <20160226154046.ga1...@quark.internal.precedence.co.uk>,
> Patrick Welche   wrote:
> >On Fri, Feb 26, 2016 at 02:30:21PM +0100, Stephan wrote:
> >> The assembly looks like junk and considering the adresses, you have
> >> tried to disassemble some memory on the stack. Is this a 64-bit wfica
> >> binary?
> >
> 
> I commented out the x86_64 restriction and it runs on amd64/current.
> At least it brings up a window saying it needs me to specify application
> servers which I don't have any...

Indeed - it runs, but once you actually start using a connected socket
recvmsg ends up returning errno 35 Resource temporarily unavailable
until the timer SEGVs it.

P

Re: Last version of of citrix Client core dumps (Was Re: How to run Microsoft Internet Explorer on NetBSD?)

[Christos Zoulas]

In article <20160226154046.ga1...@quark.internal.precedence.co.uk>,
Patrick Welche   wrote:
>On Fri, Feb 26, 2016 at 02:30:21PM +0100, Stephan wrote:
>> The assembly looks like junk and considering the adresses, you have
>> tried to disassemble some memory on the stack. Is this a 64-bit wfica
>> binary?
>

I commented out the x86_64 restriction and it runs on amd64/current.
At least it brings up a window saying it needs me to specify application
servers which I don't have any...

christos

Re: Last version of of citrix Client core dumps (Was Re: How to run Microsoft Internet Explorer on NetBSD?)

[Patrick Welche]

On Fri, Feb 26, 2016 at 02:30:21PM +0100, Stephan wrote:
> The assembly looks like junk and considering the adresses, you have
> tried to disassemble some memory on the stack. Is this a 64-bit wfica
> binary?

This is from:

/usr/pkgsrc/distfiles/linuxx64-13.3.0.344519.tar.gz

> Either way, this seems to be a completely different case. How does the
> backtrace look like?

and you're right, that was the shell script(!)

$ gdb /usr/pkg/lib/ICAClient/wfica  wfica.core 
GNU gdb (GDB) 7.10.1
Copyright (C) 2015 Free Software Foundation, Inc.
...
warning: A handler for the OS ABI "GNU/Linux" is not built into this 
configuration
of GDB.  Attempting to continue with the default i386:x86-64 settings.

Reading symbols from /usr/pkg/lib/ICAClient/wfica...(no debugging symbols 
found)...done.
[New process 26159]
[New process 6029]
[New process 18257]
[1]   Segmentation fault (core dumped) gdb /usr/pkg/lib/ICAClient/wfica 
wfica.core

$ file /usr/pkg/lib/ICAClient/wfica wfica.core
/usr/pkg/lib/ICAClient/wfica: ELF 64-bit LSB executable, x86-64, version 1 
(SYSV), dynamically linked, interpreter /lib64/ld-linux-x86-64.so.2, for 
GNU/Linux 2.6.26, BuildID[sha1]=82aabbc0e9bf9c7211ddbeb8ff679cc6fdb75d88, 
stripped
wfica.core:   ELF 64-bit LSB core file x86-64, version 1 
(SYSV), too many program headers (495)

so not much luck...

Cheers,

Patrick

Re: Last version of of citrix Client core dumps (Was Re: How to run Microsoft Internet Explorer on NetBSD?)

[Stephan]

The assembly looks like junk and considering the adresses, you have
tried to disassemble some memory on the stack. Is this a 64-bit wfica
binary?

Either way, this seems to be a completely different case. How does the
backtrace look like?

2016-02-26 13:01 GMT+01:00 Patrick Welche :
> On Fri, Feb 26, 2016 at 09:22:38AM +0100, Stephan wrote:
>> I still recommend Receiver for HTML5 in this case.
>>
>> The dump looks like a mess and eventually gdb is unable to process
>> this dump of a Linux binary on NetBSD correctly. It would be
>> interesting to know what is mapped at 0xba90004d. You could break at
>> that adress (b *0xba90004d) and check with pmap. Also, what´s the
>> corresponding instruction (x/i 0xba90004d)?
>
> Getting to stop while running for the pmap is proving interesting (need
> authentication via pnbrowse before dump in wfica).
>
> The instruction seems to always be the same though:
>
> Core was generated by `wfica'.
> Program terminated with signal SIGSEGV, Segmentation fault.
> #0  0x7f7fee73a9f4 in ?? ()
> [Current thread is 1 (process 27767)]
> (gdb) x/20i 0x7f7fee73a9e8
>0x7f7fee73a9e8:  add%al,(%rax)
>0x7f7fee73a9ea:  add%al,(%rax)
>0x7f7fee73a9ec:  add%al,(%rax)
>0x7f7fee73a9ee:  add%al,(%rax)
>0x7f7fee73a9f0:  push   %rsi
>0x7f7fee73a9f1:  rex.R
>0x7f7fee73a9f2:  rex.RB push %r13
> => 0x7f7fee73a9f4:  rex.RB
>0x7f7fee73a9f5:  rex.WRB
>0x7f7fee73a9f6:  cs rex.R
>0x7f7fee73a9f8:  rex.WR
>0x7f7fee73a9f9:  rex.WR add %r8b,(%rax)
>0x7f7fee73a9fc:  add%dl,0x75(%rcx)
>0x7f7fee73a9ff:  gs jne 0x7f7fee73aa67
>0x7f7fee73aa02:  add%al,(%rax)
>0x7f7fee73aa04:  add%al,(%rax)
>0x7f7fee73aa06:  add%al,(%rax)
>0x7f7fee73aa08:  stc
>0x7f7fee73aa09:  mov%al,%al
>0x7f7fee73aa0b:  idivl  0x7f(%rdi)
>
>
> Cheers,
>
> Patrick

Re: Last version of of citrix Client core dumps (Was Re: How to run Microsoft Internet Explorer on NetBSD?)

[Patrick Welche]

On Fri, Feb 26, 2016 at 09:22:38AM +0100, Stephan wrote:
> I still recommend Receiver for HTML5 in this case.
> 
> The dump looks like a mess and eventually gdb is unable to process
> this dump of a Linux binary on NetBSD correctly. It would be
> interesting to know what is mapped at 0xba90004d. You could break at
> that adress (b *0xba90004d) and check with pmap. Also, what´s the
> corresponding instruction (x/i 0xba90004d)?

Getting to stop while running for the pmap is proving interesting (need
authentication via pnbrowse before dump in wfica).

The instruction seems to always be the same though:

Core was generated by `wfica'.
Program terminated with signal SIGSEGV, Segmentation fault.
#0  0x7f7fee73a9f4 in ?? ()
[Current thread is 1 (process 27767)]
(gdb) x/20i 0x7f7fee73a9e8
   0x7f7fee73a9e8:  add%al,(%rax)
   0x7f7fee73a9ea:  add%al,(%rax)
   0x7f7fee73a9ec:  add%al,(%rax)
   0x7f7fee73a9ee:  add%al,(%rax)
   0x7f7fee73a9f0:  push   %rsi
   0x7f7fee73a9f1:  rex.R
   0x7f7fee73a9f2:  rex.RB push %r13
=> 0x7f7fee73a9f4:  rex.RB
   0x7f7fee73a9f5:  rex.WRB
   0x7f7fee73a9f6:  cs rex.R
   0x7f7fee73a9f8:  rex.WR
   0x7f7fee73a9f9:  rex.WR add %r8b,(%rax)
   0x7f7fee73a9fc:  add%dl,0x75(%rcx)
   0x7f7fee73a9ff:  gs jne 0x7f7fee73aa67
   0x7f7fee73aa02:  add%al,(%rax)
   0x7f7fee73aa04:  add%al,(%rax)
   0x7f7fee73aa06:  add%al,(%rax)
   0x7f7fee73aa08:  stc
   0x7f7fee73aa09:  mov%al,%al
   0x7f7fee73aa0b:  idivl  0x7f(%rdi)


Cheers,

Patrick

Re: Last version of of citrix Client core dumps (Was Re: How to run Microsoft Internet Explorer on NetBSD?)

[Stephan]

I still recommend Receiver for HTML5 in this case.

The dump looks like a mess and eventually gdb is unable to process
this dump of a Linux binary on NetBSD correctly. It would be
interesting to know what is mapped at 0xba90004d. You could break at
that adress (b *0xba90004d) and check with pmap. Also, what´s the
corresponding instruction (x/i 0xba90004d)?

2016-02-25 15:42 GMT+01:00 Patrick Welche :
> On Thu, Feb 25, 2016 at 02:44:32PM +0100, Stephan wrote:
>> Why don´t you just use the Receiver for HTML5? With regard to your
>> crash, do you have a backtrace handy?
>
> Have a look at Jose's from earlier in this thread:
>
> http://mail-index.netbsd.org/netbsd-users/2016/02/03/msg017788.html
>
> P

Re: Last version of of citrix Client core dumps (Was Re: How to run Microsoft Internet Explorer on NetBSD?)

[Patrick Welche]

On Thu, Feb 25, 2016 at 02:44:32PM +0100, Stephan wrote:
> Why don´t you just use the Receiver for HTML5? With regard to your
> crash, do you have a backtrace handy?

Have a look at Jose's from earlier in this thread:

http://mail-index.netbsd.org/netbsd-users/2016/02/03/msg017788.html

P

Re: Last version of of citrix Client core dumps (Was Re: How to run Microsoft Internet Explorer on NetBSD?)

[Stephan]

Why don´t you just use the Receiver for HTML5? With regard to your
crash, do you have a backtrace handy?

2016-02-25 14:03 GMT+01:00 Patrick Welche :
> On Wed, Feb 03, 2016 at 07:31:35PM +0100, Jose Luis Rodriguez Garcia wrote:
>> On Tue, Feb 2, 2016 at 4:17 AM, Eric Haszlakiewicz  wrote:
>> > On 2/1/2016 3:51 PM, Jose Luis Rodriguez Garcia wrote:
>>
>>
>> > Have you tried downloading a newer version of the client from Citrix's 
>> > site?
>>
>> I have just tried the last version of citrix as you told me.
>>
>> It solves the problem of the certificate (the old version didn't
>> understand the CA from Godaddy, but it coredumps at the startup.
>>
>> I copy backtrace/registers and output from ktruss:
> ...
>>   2412   2412 wficaRET   writev 324/0x144
>>   2412   2412 wficaCALL  poll(0xbfbff418,1,0x)
>>   2412   2412 wficaRET   poll 1
>>   2412   2412 wficaCALL  read(3,0x8264a38,0x1000)
>>   2412   2412 wficaGIO   fd 3 read 32 bytes
>>"\^A\^B8\^B\0\0\0\0002\0`\^A\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"
>>   2412   2412 wficaRET   read 32/0x20
>>   2412   2412 wficaCALL  read(3,0x8264a38,0x1000)
>>   2412   2412 wficaRET   read -1 unknown errno 35
>>   2412   2412 wficaCALL  read(3,0x8264a38,0x1000)
>>   2412   2412 wficaRET   read -1 unknown errno 35
>>   2412   2412 wficaCALL  gettimeofday(0xbfbff8bc,0)
>>   2412   2412 wficaRET   gettimeofday 0
>>   2412   2412 wficaCALL  gettimeofday(0xbfbff83c,0)
>>   2412   2412 wficaRET   gettimeofday 0
>>   2412   2412 wficaCALL  gettimeofday(0xbfbff7dc,0)
>>   2412   2412 wficaRET   gettimeofday 0
>>   2412   2412 wficaCALL  socketcall(9,0xbfbff5f0)
>>   2412   2412 wficaMISC  send: 16,
>> 0600ffe87cffdcffb92200
>>   2412   2412 wficaMISC  msghdr: [name=0x0, namelen=0,
>> iov=0xf323bc84, iovlen=1, control=0x0, controllen=3226819742, flags=0]
>>   2412   2412 wficaGIO   fd 6 wrote 34 bytes
>>
>> "\^W\^C\^A\0\^]\\\M-f`WIH*\M^F\M-%\^X\M-y\M^@9\M-t({In39$\M-!\^[\M^TNc\M^N/\M-_"
>>   2412   2412 wficaRET   socketcall 34/0x22
>>   2412   2412 wficaCALL  gettimeofday(0xbfbff6fc,0)
>>   2412   2412 wficaRET   gettimeofday 0
>>   2412   2412 wficaCALL  gettimeofday(0xbfbff68c,0)
>>   2412   2412 wficaRET   gettimeofday 0
>>   2412   2412 wficaCALL  gettimeofday(0xbfbff6fc,0)
>>   2412   2412 wficaRET   gettimeofday 0
>>   2412   2412 wficaPSIG  SIGSEGV SIG_DFL: code=SEGV_MAPERR,
>> addr=0x92ee48c6, trap=14)
>>   2412   2459 wficaRET   select -1 unknown errno 4
>>   2412   2412 wficaNAMI  "wfica.core"
>
> I am seeing this too. Essentially there are a load of calls to
> linux_sys_recvmsg which returns 35 (EAGAIN). The last round looks like
>
>  got here linux_sys_recvmsg 834 (do_sys_recvmsg=35)
>  got here linux_select1 894 selcommon=0
>  got here linux_sys_select 844
>  got here linux_sys_recvmsg 815
>  got here linux_to_bsd_msghdr 455
>  got here linux_sys_recvmsg 820
>  got here linux_to_bsd_msg_flags 279
>  msg_name=0x0
>  msg_namelen=0
>  msg_iov=0x7f7febd0
>  msg_iov=0x7f
>  msg_control=0x7f7fec20
>  msg_controllen=80
>  msg_flags=400
>  control=0x80586483
>  from=0xfe804534be58
>  got here linux_sys_recvmsg 834 (do_sys_recvmsg=35)
>  got here linux_select1 894 selcommon=4
>  pid 3382 (wfica), uid 1000: exited on signal 11 (core dumped)
>
> So it seems that after too many retries, a timer fires and kills the
> process (linux_select1()). I suppose the underlying issue is with
> linux_sys_recvmsg, but how can you find out what?
>
> Cheers,
>
> Patrick

Re: Last version of of citrix Client core dumps (Was Re: How to run Microsoft Internet Explorer on NetBSD?)

[Patrick Welche]

On Wed, Feb 03, 2016 at 07:31:35PM +0100, Jose Luis Rodriguez Garcia wrote:
> On Tue, Feb 2, 2016 at 4:17 AM, Eric Haszlakiewicz  wrote:
> > On 2/1/2016 3:51 PM, Jose Luis Rodriguez Garcia wrote:
> 
> 
> > Have you tried downloading a newer version of the client from Citrix's site?
> 
> I have just tried the last version of citrix as you told me.
> 
> It solves the problem of the certificate (the old version didn't
> understand the CA from Godaddy, but it coredumps at the startup.
> 
> I copy backtrace/registers and output from ktruss:
...
>   2412   2412 wficaRET   writev 324/0x144
>   2412   2412 wficaCALL  poll(0xbfbff418,1,0x)
>   2412   2412 wficaRET   poll 1
>   2412   2412 wficaCALL  read(3,0x8264a38,0x1000)
>   2412   2412 wficaGIO   fd 3 read 32 bytes
>"\^A\^B8\^B\0\0\0\0002\0`\^A\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"
>   2412   2412 wficaRET   read 32/0x20
>   2412   2412 wficaCALL  read(3,0x8264a38,0x1000)
>   2412   2412 wficaRET   read -1 unknown errno 35
>   2412   2412 wficaCALL  read(3,0x8264a38,0x1000)
>   2412   2412 wficaRET   read -1 unknown errno 35
>   2412   2412 wficaCALL  gettimeofday(0xbfbff8bc,0)
>   2412   2412 wficaRET   gettimeofday 0
>   2412   2412 wficaCALL  gettimeofday(0xbfbff83c,0)
>   2412   2412 wficaRET   gettimeofday 0
>   2412   2412 wficaCALL  gettimeofday(0xbfbff7dc,0)
>   2412   2412 wficaRET   gettimeofday 0
>   2412   2412 wficaCALL  socketcall(9,0xbfbff5f0)
>   2412   2412 wficaMISC  send: 16,
> 0600ffe87cffdcffb92200
>   2412   2412 wficaMISC  msghdr: [name=0x0, namelen=0,
> iov=0xf323bc84, iovlen=1, control=0x0, controllen=3226819742, flags=0]
>   2412   2412 wficaGIO   fd 6 wrote 34 bytes
>
> "\^W\^C\^A\0\^]\\\M-f`WIH*\M^F\M-%\^X\M-y\M^@9\M-t({In39$\M-!\^[\M^TNc\M^N/\M-_"
>   2412   2412 wficaRET   socketcall 34/0x22
>   2412   2412 wficaCALL  gettimeofday(0xbfbff6fc,0)
>   2412   2412 wficaRET   gettimeofday 0
>   2412   2412 wficaCALL  gettimeofday(0xbfbff68c,0)
>   2412   2412 wficaRET   gettimeofday 0
>   2412   2412 wficaCALL  gettimeofday(0xbfbff6fc,0)
>   2412   2412 wficaRET   gettimeofday 0
>   2412   2412 wficaPSIG  SIGSEGV SIG_DFL: code=SEGV_MAPERR,
> addr=0x92ee48c6, trap=14)
>   2412   2459 wficaRET   select -1 unknown errno 4
>   2412   2412 wficaNAMI  "wfica.core"

I am seeing this too. Essentially there are a load of calls to
linux_sys_recvmsg which returns 35 (EAGAIN). The last round looks like

 got here linux_sys_recvmsg 834 (do_sys_recvmsg=35)
 got here linux_select1 894 selcommon=0
 got here linux_sys_select 844
 got here linux_sys_recvmsg 815
 got here linux_to_bsd_msghdr 455
 got here linux_sys_recvmsg 820
 got here linux_to_bsd_msg_flags 279
 msg_name=0x0
 msg_namelen=0
 msg_iov=0x7f7febd0
 msg_iov=0x7f
 msg_control=0x7f7fec20
 msg_controllen=80
 msg_flags=400
 control=0x80586483
 from=0xfe804534be58
 got here linux_sys_recvmsg 834 (do_sys_recvmsg=35)
 got here linux_select1 894 selcommon=4
 pid 3382 (wfica), uid 1000: exited on signal 11 (core dumped)

So it seems that after too many retries, a timer fires and kills the
process (linux_select1()). I suppose the underlying issue is with
linux_sys_recvmsg, but how can you find out what?

Cheers,

Patrick

Last version of of citrix Client core dumps (Was Re: How to run Microsoft Internet Explorer on NetBSD?)

[Jose Luis Rodriguez Garcia]

On Tue, Feb 2, 2016 at 4:17 AM, Eric Haszlakiewicz  wrote:
> On 2/1/2016 3:51 PM, Jose Luis Rodriguez Garcia wrote:


> Have you tried downloading a newer version of the client from Citrix's site?

I have just tried the last version of citrix as you told me.

It solves the problem of the certificate (the old version didn't
understand the CA from Godaddy, but it coredumps at the startup.

I copy backtrace/registers and output from ktruss:


pc1$ gdb wfica wfica.core
warning: A handler for the OS ABI "GNU/Linux" is not built into this
configuration
of GDB.  Attempting to continue with the default i386 settings.

Reading symbols from /home/jose/citrix/wfica...(no debugging symbols
found)...done.
[New process 2412]
[New process 2227]
[New process 2459]

warning: Can not parse XML library list; XML support was disabled at
compile time
Core was generated by `wfica'.
Program terminated with signal 11, Segmentation fault.
#0  0xba90004d in ?? ()
(gdb) info registers
eax0x80bacba134982842
ecx0x82f7f60137330528
edx0xc12
ebx0x2032
esp0xbfbff6e00xbfbff6e0
ebp0xbfbff7080xbfbff708
esi0x8214c04136399876
edi0xba90004c-1164967860
eip0xba90004d0xba90004d
eflags 0x10202[ IF RF ]
cs 0xa3163
ss 0x2335
ds 0x2335
es 0x2335
fs 0x2335
gs 0xb3179
(gdb) bt
#0  0xba90004d in ?? ()
#1  0x080bfd57 in EmulProcessInput ()
#2  0xbaa51e90 in ?? ()
#3  0x08213720 in ?? ()
#4  0x080ca9c0 in ?? ()
#5  0x080cdd97 in ?? ()
#6  0x080caa3b in ?? ()
#7  0xbaa513b2 in ?? ()
#8  0x08214f00 in ?? ()
#9  0x080c91e9 in WdPoll ()
#10 0x0813e92d in Call_WD ()
#11 0x08104144 in Do_work ()
#12 0x08103cf5 in ?? ()
#13 0xbb2bb4ff in ?? ()
#14 0x in ?? ()
(gdb) thread apply all backtrace

Thread 3 (process 2459):
#0  0xbb1ba907 in ?? ()
#1  0x041e in ?? ()
#2  0x0013 in ?? ()
#3  0x0820e5f8 in ?? ()
#4  0x08084013 in ?? ()
#5  0xbb25fa7d in ?? ()
#6  0x0820e5f8 in ?? ()
#7  0xbb1c1abe in ?? ()
Backtrace stopped: Not enough registers or memory available to unwind further

Thread 2 (process 2227):
#0  0xbb25fb24 in ?? ()
#1  0xbaa79000 in ?? ()
#2  0x001fb000 in ?? ()
#3  0x0004 in ?? ()
#4  0x in ?? ()

Thread 1 (process 2412):
#0  0xba90004d in ?? ()
#1  0x080bfd57 in EmulProcessInput ()
#2  0xbaa51e90 in ?? ()
#3  0x08213720 in ?? ()
#4  0x080ca9c0 in ?? ()
#5  0x080cdd97 in ?? ()
#6  0x080caa3b in ?? ()
#7  0xbaa513b2 in ?? ()
#8  0x08214f00 in ?? ()
#9  0x080c91e9 in WdPoll ()
#10 0x0813e92d in Call_WD ()
#11 0x08104144 in Do_work ()
#12 0x08103cf5 in ?? ()
#13 0xbb2bb4ff in ?? ()
#14 0x in ?? ()
(gdb)

Ktruss output:

  2412   2412 wficaGIO   fd 3 wrote 324 bytes
   ";\^C\^E\0=\0 \^B\0\0\0\09\^C\"\^B\M^C\0\r\08\^C\^D\0=\0
\^B\b\0\0\0\0\0\0\0H\0G\0007\0 \^B=\0
\^B\M^G\0\r\09\^C\"\^B\0\^A\^A\0\M^?\M^?\M^?\M^?\

\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M-3\M-/\0\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M-3\M-/\0\^?\

\M^?\M^?\M-~\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M-{\M^?\M^?\M-o\M^?\M^?\M-3\M-/\0C\M^?\M^?\M-^\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M-{\M^?\M-n\M-o\M^?\M^?\M-3\

\M-/\0}\M^?\M^?\M-~\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M-{\M^?\M-~\M-o\M^?\M^?\M-3\M-/\0~8\M^NV8\M-<\^[\M-;\M-z\M-(q\^X\M-Chl\M-|\M^?\M-3\M-/\0~\M-W\

\M-uZ\M-W\M-=\M-k\M-:|\M-O\M-.\M-k\M-:\M-n\M-+\M^?\M^?\M-3\M-/\0~\^W\M-t\\\M-W}\M-m\M-:~\M-o\240\v\M-:n(\M^?\M^?\M-3\M-/\0~\M-W\M-wZ\M-W}\M-m\

\M-:~\M-o\M->\M-k\M-;\M-.\M-k\M-|\M^?\M-3\M-/\0}\M-W\M-uV\M-W\M-}\M-n\M-:~\M-o\M-.\M-k\M-:\M-.\M-k\M-m\M-n\M-3\M-/\0C7\M^NN7\M-|\^^\M^G\M-~\M-h\

q\^X\M-;i(\M-n\M-n\M-3\M-/\0\M^?\M^?\M^?\M^?\M^?}\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M^?\M-2\M--\0\M^?\M^?\M^?\M^??~\M^?\M^?\M^?\M^?\M^?\
\M^?\M^?\M^?\M^?\M^?\M^?\M-/\M-*\0+\M-.\^A\0"
  2412   2412 wficaRET   writev 324/0x144
  2412   2412 wficaCALL  poll(0xbfbff418,1,0x)
  2412   2412 wficaRET   poll 1
  2412   2412 wficaCALL  read(3,0x8264a38,0x1000)
  2412   2412 wficaGIO   fd 3 read 32 bytes
   "\^A\^B8\^B\0\0\0\0002\0`\^A\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0\0"
  2412   2412 wficaRET   read 32/0x20
  2412   2412 wficaCALL  read(3,0x8264a38,0x1000)
  2412   2412 wficaRET   read -1 unknown errno 35
  2412   2412 wficaCALL  read(3,0x8264a38,0x1000)
  2412   2412 wficaRET   read -1 unknown errno 35
  2412   2412 wficaCALL  gettimeofday(0xbfbff8bc,0)
  2412   2412 wficaRET   gettimeofday 0
  2412   2412 wficaCALL  gettimeofday(0xbfbff83c,0)
  2412   2412 wficaRET   gettimeofday 0
  2412   2412 wficaCALL  gettimeofday(0xbfbff7dc,0)
  2412   2412 wficaRET   gettimeofday 0
  2412   2412 wficaCALL  socketcall(9,0xbfbff5f0)
  2412   2412 wficaMISC  send: 16,
0600ffe87cffdcffb92200
  

Re: How to run Microsoft Internet Explorer on NetBSD?

[Jose Luis Rodriguez Garcia]

> If you're content with VirtualBox, Microsoft themselves provide images
> of various versions of Windows with various versions of IE already
> installed.
>
> https://dev.windows.com/en-us/microsoft-edge/tools/vms/windows/

On Fri, Jan 29, 2016 at 2:08 PM, Marina Brown  wrote:

> They have a portal that requires windows. I think you can even get
> windows 10 for free under some circumstances. Not sure about this i had
> a windows license owned by my company.
>
> --- Marina Brown

Thank Marina, Coypu

I have opted by this way.

I have downloaded the image for virtual box for ie6_winxp. It has
worked with the qemu emulator.

I have done:

tar xf file.ovf
It has uncompressed a vmdk file. I have converted it to qcow2 with qemu-img:
qemu-img -c -O qcow2 "IE6 - WinXP-disk1.vmdk" "IE6 - WinXP-disk1.qcow2"

qemu-system-i386 -m 256 -net nic,model=ne2k_pci  -net user -hda "IE6 -
WinXP-disk1.qcow2"

and it worked ok.

It is slow..., but it works and it is easy if you think on wine.

Re: How to run Microsoft Internet Explorer on NetBSD?

[Jose Luis Rodriguez Garcia]

>The so-called "Citrix Receiver" ? I've tried it using both Wine and Linux 
>emulation. The Linux version was a huge pain. It segfaulted, whined about SSL 
>/ x509 certificates (so tired of seeing this lately in apps), and had the 
>usual way-too-many-dependencies on a zillion worthless GUI libraries and 
>abstraction layers. Honestly, the Wine version looked like it was going to 
>work up until the last moment when ... it didn't. With Windows apps you can't 
>do much but just shrug when they fail.

I had always the same problem. I have read that the https problem is
resolved installing the crypt32 package (winetricks). But I have been
unable to test it. Explorer 7 and Explorer 8 fails to install.

Explorer 6 installs sucessfully (with crypt32) but crashes with the
SSL problem that you mention.

I  have read some report that may be for Explorer 8 it can help wine
mono/wine gecko native. May be I can try to compile them.

> It seems like the previous ICA client didn't have nearly as much kruft and 
> actually worked. I know I've used it at some point on NetBSD. Of course, that 
> version is probably no longer around. It's too bad, though. IIRC, you could 
> often use older clients with newer Citrix servers.

It is still in pkgsrc : net/citrix_ica version 10.6.115659.

It worked until my company changed the certificates to godaddy. I have
been unable to configure the SSL certificates. I am thinking that it
can be the SSL client doesn't understand the new certificates. Can it
be?

-Swift

On Fri, Jan 29, 2016 at 11:31 AM, Jose Luis Rodriguez Garcia
 wrote:
> I need to use Internet Explorer for access a web from my company:
> Siebel software.
>
> It uses Active X and it only works with Internet Explorer. (They
> haven't activated the mode of Siebel Software for non Microsoft
> browsers).
>
> .It is a application that I must use, very few times.
>
> It is the option that I have used (tested some of them). What option
> do you suggest?
>
> 1- Wine (I have tried it years ago with bad results with Explorer.
> 2- I can access with Citrix, but the citrix client from pkgsrc is very
> old ,and it has problems with certificates of our Citrix Server. I
> have tried to add the certificates without success. I don't know if I
> am doing the right thing, or the citrix client has some problem for
> understand the new certificates.
> 3- VMware.
> 4- Other emulator?

Re: How to run Microsoft Internet Explorer on NetBSD?

[Hal Murray]

swiftgri...@gmail.com said:
> hat 99% of folks who use SSL care about is  _transport_ encryption, NOT the
> chain-of-trust, which I consider to be  fundamentally flawed and broken at
> it's very core.

Without something like a chain-of-trust you don't know that your encrypted 
connection is going to the right site.

A man-in-the-middle can claim to be your bank.  How do you propose verify 
that?

-- 
These are my opinions.  I hate spam.

SSL makes me crazy (was Re: How to run Microsoft Internet Explorer on NetBSD?)

[Swift Griggs]

On Mon, 1 Feb 2016, Hal Murray wrote:

Without something like a chain-of-trust you don't know that your encrypted
connection is going to the right site.


I understand it's design purpose, but I disagree with where the design 
puts that trust. When it comes down to brass-tacks, do you trust Verisign 
is doing what they say they do to verify that the cert holder is the party 
you want to have an encrypted conversation with ? My answer to that 
question is "hell no". I don't trust Verisign or any other corporation 
that would be a CA under our current system. Thus, I think the system is 
flawed.


A man-in-the-middle can claim to be your bank.  How do you propose 
verify that?


Well, the way I understand it, (and I'm probably wrong) but a 
man-in-the-middle would have to be able to break Diffie Hellman unless you 
can force a key update. It doesn't have much to do with the cert being 
presented.  So, I'm not sure that's true (not trying to be difficult or 
troll, just saying). However, I do take your point. Ie.. how do you verify 
the remote party's identity without a trusted 3rd party saying "Yeah, 
that's him" ?  My preferred answer would involve removing the trust from 
the dirtbag corporation and giving it to another entity. Some 
possibilities include:


* A non-profit organization with fewer motives to get in bed with the NSA
  or other corporations.

* A pool or group of trusted users who rate / rank trustability. People
  with a vested interest in getting it right and difficult to pay off or
  bribe.

* Get rid of the trust idea altogether and use some kind of
  physical or manual challenge-response. The genius would be in coming up
  with one simple enough to work, yet maintain security. Do you really
  think folks are clicking on the cert and following the chain of trust
  anyway ? Most users don't even understand it's happening (not good).

I'm not saying that the same issue (authentication of a remote party's 
identity) wouldn't come up in any system you created. However, I am saying 
that SSL has done an exceptionally poor job at...  well... it's job. It's 
over-complicated, apparently quite insecure. So insecure in fact that it's 
been nearly completely broken twice. Each time the fixes have been 
increasingly painful and disruptive enough to warrant asking the question: 
Is SSL really a good system? My experience as a user and admin would 
prompt me to answer "No way, Jose. Start again without the committee."


As an example, PGP was designed well before SSL. PGP has survived all this 
time without any exposures on the order of what we've seen with SSL (it's 
had plenty of coding issues, but no completely-busted algorithm issues). 
It's also quite a bit more simple (and that's kind of my point). 
Complexity is the enemy of security since it only provides more attack 
surface. I would submit that to "secure" is most of the time to simplify.


It's nothing personal against you, Hal, or anyone else. Hopefully, nobody 
here used to work for Netscape or other folks involved with designing SSL. 
I just think SSL was badly designed from the start and I believe the facts 
(the security issues) back me up.


-Swift

Re: How to run Microsoft Internet Explorer on NetBSD?

[Swift Griggs]

On Mon, 1 Feb 2016, Jose Luis Rodriguez Garcia wrote:

It is still in pkgsrc : net/citrix_ica version 10.6.115659.


Ugh. I forgot about that. I need to go back to i386. It fails for AMD64, 
but yeah, it's still a certificate trust-nightmare.


 I used to mildly dislike SSL before it was completely broken by the 
NSA and others. Now, I consider it WAY too complicated to believe in and I 
really hate it. This whole idiocy with "trust" has gone down some road 
where I'm supposed to trust that one dirty corporation (like Verisign) 
believes that another slimey corporation is legit, (and their verification 
steps are easily defrauded/circumvented). What if I fundamentally despise 
both of them and don't trust either one? Then all I am left with is a 
turd-hunt for the latest certificate bundle that XYZ crapware needs to run 
and a bad attitude. What 99% of folks who use SSL care about is 
_transport_ encryption, NOT the chain-of-trust, which I consider to be 
fundamentally flawed and broken at it's very core. However, it's the 
chain-of-trust features that drag down SSL the most. Encryption is hard 
enough to understand & manage without adding in a double-batch of 
committee-based stupidity. Okay, now I'll be over here crying "KISS!" in 
the wilderness if anyone needs me...  :-P 


Also, IIRC, even on i386 (from memory, because I fiddled with it about 6-8 
months ago) it had major issues. The main thing I ran into was that 
without the browser plugin you just get prompted for what to do with an 
".ICA" file when the Citrix portal throws that at you. The package 
compiled & installed, but once I actually tried to use it, I noticed it 
had some other major issues (and it seems like they were more than just 
certificate issues). In fact, IIRC, I think the 'wfica' binary was missing 
some libraries it was linked to, and that was my biggest problem. Check 
your output from: ldd `which wfica`


Thanks,
  Swift

Re: SSL makes me crazy (was Re: How to run Microsoft Internet Explorer on NetBSD?)

[Hal Murray]

swiftgri...@gmail.com said:
> Well, the way I understand it, (and I'm probably wrong) but a
> man-in-the-middle would have to be able to break Diffie Hellman

How did you get your banks public key?  Without a chain-of-trust you have to 
get it on your own and the man in the middle has a good chance of subverting 
that process.

You aren't the only one who dislikes the current system, but nobody has come up 
with a better plan.  Yet.


> When it comes down to brass-tacks, do you trust Verisign is doing what they
> say they do to verify that the cert holder is the party  you want to have an
> encrypted conversation with ?

Verisign has serious incentives to do the right thing.  If they screwup they 
are likely to go out of business.  The NSA may be able to twist their arm, but 
Verizon or Comcast probably can't.

I think the major certificate issuing companies have various degrees of 
checking.  I don't know the details.  Checking costs money.  I think some of 
the options are serious enough to be appropriate for banks.

If I was going to put serious effort into this area, I'd look into a UI to 
display the chain so I could get convenient reminders about who was signing 
things I used.  Maybe a nightly summary.  Maybe a confirm step if the top level 
signer was strange where I get to maintain a white-list of sites that are 
non-strange for me.


-- 
These are my opinions.  I hate spam.

Re: SSL makes me crazy (was Re: How to run Microsoft Internet Explorer on NetBSD?)

[Miguel C]

On Mon, Feb 1, 2016 at 11:16 PM, Swift Griggs  wrote:

> On Mon, 1 Feb 2016, Hal Murray wrote:
>
>> Without something like a chain-of-trust you don't know that your encrypted
>> connection is going to the right site.
>>
>
> I understand it's design purpose, but I disagree with where the design
> puts that trust. When it comes down to brass-tacks, do you trust Verisign
> is doing what they say they do to verify that the cert holder is the party
> you want to have an encrypted conversation with ? My answer to that
> question is "hell no". I don't trust Verisign or any other corporation that
> would be a CA under our current system. Thus, I think the system is flawed.
>
> A man-in-the-middle can claim to be your bank.  How do you propose verify
>> that?
>>
>
> Well, the way I understand it, (and I'm probably wrong) but a
> man-in-the-middle would have to be able to break Diffie Hellman unless you
> can force a key update. It doesn't have much to do with the cert being
> presented.  So, I'm not sure that's true (not trying to be difficult or
> troll, just saying). However, I do take your point. Ie.. how do you verify
> the remote party's identity without a trusted 3rd party saying "Yeah,
> that's him" ?  My preferred answer would involve removing the trust from
> the dirtbag corporation and giving it to another entity. Some possibilities
> include:
>
> * A non-profit organization with fewer motives to get in bed with the NSA
>   or other corporations.
>
> * A pool or group of trusted users who rate / rank trustability. People
>   with a vested interest in getting it right and difficult to pay off or
>   bribe.
>
> * Get rid of the trust idea altogether and use some kind of
>   physical or manual challenge-response. The genius would be in coming up
>   with one simple enough to work, yet maintain security. Do you really
>   think folks are clicking on the cert and following the chain of trust
>   anyway ? Most users don't even understand it's happening (not good).
>
> I'm not saying that the same issue (authentication of a remote party's
> identity) wouldn't come up in any system you created. However, I am saying
> that SSL has done an exceptionally poor job at...  well... it's job. It's
> over-complicated, apparently quite insecure. So insecure in fact that it's
> been nearly completely broken twice. Each time the fixes have been
> increasingly painful and disruptive enough to warrant asking the question:
> Is SSL really a good system? My experience as a user and admin would prompt
> me to answer "No way, Jose. Start again without the committee."
>
> As an example, PGP was designed well before SSL. PGP has survived all this
> time without any exposures on the order of what we've seen with SSL (it's
> had plenty of coding issues, but no completely-busted algorithm issues).
> It's also quite a bit more simple (and that's kind of my point). Complexity
> is the enemy of security since it only provides more attack surface. I
> would submit that to "secure" is most of the time to simplify.
>
> It's nothing personal against you, Hal, or anyone else. Hopefully, nobody
> here used to work for Netscape or other folks involved with designing SSL.
> I just think SSL was badly designed from the start and I believe the facts
> (the security issues) back me up.
>
> -Swift
>

Sorry to bum in, but are you aware of --> https://letsencrypt.org/ !?

Re: How to run Microsoft Internet Explorer on NetBSD?

[Eric Haszlakiewicz]

On 2/1/2016 3:51 PM, Jose Luis Rodriguez Garcia wrote:
It is still in pkgsrc : net/citrix_ica version 10.6.115659. It worked 
until my company changed the certificates to godaddy. I have been 
unable to configure the SSL certificates. I am thinking that it can be 
the SSL client doesn't understand the new certificates. Can it be? 
Sure, I supposed it's possible.  That is a 9 year old package and lots 
could have changed in the meanwhile.
Were you able to get the new certificate in the right format? (looks 
like the files in /usr/pkg/lib/ICAClient/keystore/cacerts are in DER format)
Do you happen to know what's different about your company's old and new 
certs?  If you look at the details of the cert (e.g. with "openssl x509 
-inform DER -in foo.crt -text") does anything jump out at you as being 
different?


Have you tried downloading a newer version of the client from Citrix's 
site?  I just tried running it, and after the following steps it at 
least seems to start:

  pkgin install suse_locale suse_motif suse_x11
  pkgin install suse_openmotif-10.0nb2
  pkgin install suse_gtk2
  mkdir citrix-ica
  cd citrix-ica
  ar x icaclient_13.1.0.285639_i386.deb
  tar xzf data.tar.gz
  cd opt/Citrix/ICAClient
  export ICAROOT=`pwd`
  ./wfica

I expect that once I get an X server up and running on this machine 
it'll complain about not being able to find the .../etc files, but I 
suspect that might not be too hard to fix (i.e. drop them in 
/emul/linux/etc)


Eric

Re: How to run Microsoft Internet Explorer on NetBSD?

[David Brownlee]

On 30 January 2016 at 20:26, Jose Luis Rodriguez Garcia
 wrote:
> I have tried this other time. (I tried it in the past also without success).
>
> The package is broken, and one file that tries to download from
> Microsoft is not longer available. (mfc42.cab)
>
> After of downloading manually the missing package from web archiv,
> fixing the scripts of ies4linux, it installs Internet Explorer 6.
>
> As always (as mi previous attempts) it fails at the few seconds of
> start the explorer.
>
> Have you tried with some older version of wine that worked better than
> the actual versions?

A few years back I ran up three XP VMs under qemu for occasional use
of IE 6, 7 & 8, and still have them, though I've been fortunate enough
not to need them very much recently. Its not as nice as being able to
have a free floating browser window, and not as performant as a
VirtualBox or xen solution, but its reliable and good enough for my
needs

Re: How to run Microsoft Internet Explorer on NetBSD?

[Swift Griggs]

On Sat, 30 Jan 2016, Jose Luis Rodriguez Garcia wrote:
Does anyone tried a new version of citrix (not the one from pkgsrc) 
client in NetBSD?


The so-called "Citrix Receiver" ? I've tried it using both Wine and Linux 
emulation. The Linux version was a huge pain. It segfaulted, whined about 
SSL / x509 certificates (so tired of seeing this lately in apps), and had 
the usual way-too-many-dependencies on a zillion worthless GUI libraries 
and abstraction layers. Honestly, the Wine version looked like it was 
going to work up until the last moment when ... it didn't. With Windows 
apps you can't do much but just shrug when they fail.


It seems like the previous ICA client didn't have nearly as much kruft and 
actually worked. I know I've used it at some point on NetBSD. Of course, 
that version is probably no longer around. It's too bad, though. IIRC, you 
could often use older clients with newer Citrix servers.


-Swift

Re: How to run Microsoft Internet Explorer on NetBSD?

[Jose Luis Rodriguez Garcia]

I have tried this other time. (I tried it in the past also without success).

The package is broken, and one file that tries to download from
Microsoft is not longer available. (mfc42.cab)

After of downloading manually the missing package from web archiv,
fixing the scripts of ies4linux, it installs Internet Explorer 6.

As always (as mi previous attempts) it fails at the few seconds of
start the explorer.

Have you tried with some older version of wine that worked better than
the actual versions?


On Fri, Jan 29, 2016 at 11:55 AM, Mayuresh  wrote:
> On Fri, Jan 29, 2016 at 11:31:17AM +0100, Jose Luis Rodriguez Garcia wrote:
>> I need to use Internet Explorer for access a web from my company:
>
> www/ies4linux? [ I haven't used of late. ]
>
> Mayuresh.

Re: How to run Microsoft Internet Explorer on NetBSD?

[Jose Luis Rodriguez Garcia]

Thank you. If I have the energy, I will ask in some Citrix forum, if
they can resolve the problem with certificates.

Does anyone tried a new version of citrix (not the one from pkgsrc)
client in NetBSD?

On Fri, Jan 29, 2016 at 4:14 PM, Eric Haszlakiewicz  wrote:
> On January 29, 2016 5:31:17 AM EST, Jose Luis Rodriguez Garcia 
>  wrote:
>>I need to use Internet Explorer for access a web from my company:
>>Siebel software.
>>
>>It uses Active X and it only works with Internet Explorer. (They
>>haven't activated the mode of Siebel Software for non Microsoft
>>browsers).
>>
>>.It is a application that I must use, very few times.
>>
>>It is the option that I have used (tested some of them). What option
>>do you suggest?
>>
>>2- I can access with Citrix, but the citrix client from pkgsrc is very
>>old ,and it has problems with certificates of our Citrix Server. I
>>have tried to add the certificates without success. I don't know if I
>>am doing the right thing, or the citrix client has some problem for
>>understand the new certificates.
>
> If you're running this from inside your company's network, Citrix is probably 
> your best bet for minimizing friction with your IT/security department, and 
> making things easy for yourself.  If you can get it the Citrix client working 
> reliably then you can let someone else worry about Windows licensing, making 
> sure the right version of IE is installed, security patches are installed, 
> etc...
>
> Eric
>
>

Re: How to run Microsoft Internet Explorer on NetBSD?

[coypu]

If you're content with VirtualBox, Microsoft themselves provide images
of various versions of Windows with various versions of IE already
installed.

https://dev.windows.com/en-us/microsoft-edge/tools/vms/windows/

Re: How to run Microsoft Internet Explorer on NetBSD?

[Eric Haszlakiewicz]

On January 29, 2016 5:31:17 AM EST, Jose Luis Rodriguez Garcia 
 wrote:
>I need to use Internet Explorer for access a web from my company:
>Siebel software.
>
>It uses Active X and it only works with Internet Explorer. (They
>haven't activated the mode of Siebel Software for non Microsoft
>browsers).
>
>.It is a application that I must use, very few times.
>
>It is the option that I have used (tested some of them). What option
>do you suggest?
>
>2- I can access with Citrix, but the citrix client from pkgsrc is very
>old ,and it has problems with certificates of our Citrix Server. I
>have tried to add the certificates without success. I don't know if I
>am doing the right thing, or the citrix client has some problem for
>understand the new certificates.

If you're running this from inside your company's network, Citrix is probably 
your best bet for minimizing friction with your IT/security department, and 
making things easy for yourself.  If you can get it the Citrix client working 
reliably then you can let someone else worry about Windows licensing, making 
sure the right version of IE is installed, security patches are installed, 
etc...

Eric

Re: How to run Microsoft Internet Explorer on NetBSD?

[Marina Brown]

On 01/29/2016 05:31 AM, Jose Luis Rodriguez Garcia wrote:
> I need to use Internet Explorer for access a web from my company:
> Siebel software.
> 
> It uses Active X and it only works with Internet Explorer. (They
> haven't activated the mode of Siebel Software for non Microsoft
> browsers).
> 
> .It is a application that I must use, very few times.
> 
> It is the option that I have used (tested some of them). What option
> do you suggest?
> 
> 1- Wine (I have tried it years ago with bad results with Explorer.
> 2- I can access with Citrix, but the citrix client from pkgsrc is very
> old ,and it has problems with certificates of our Citrix Server. I
> have tried to add the certificates without success. I don't know if I
> am doing the right thing, or the citrix client has some problem for
> understand the new certificates.
> 3- VMware.
> 4- Other emulator?
> 

I use Qemu for virtualization. Not much different that VMware but it is
open source if that matters in this instance. There is virtualbox that
you can use for virtualization too. If i were you i would keep a small
windows virt for use with your company - i used to do just the same.
They have a portal that requires windows. I think you can even get
windows 10 for free under some circumstances. Not sure about this i had
a windows license owned by my company.

--- Marina Brown

Re: How to run Microsoft Internet Explorer on NetBSD?

[Mayuresh]

On Fri, Jan 29, 2016 at 11:31:17AM +0100, Jose Luis Rodriguez Garcia wrote:
> I need to use Internet Explorer for access a web from my company:

www/ies4linux? [ I haven't used of late. ]

Mayuresh.

Re: How to run Microsoft Internet Explorer on NetBSD?

[Swift Griggs]

On Fri, 29 Jan 2016, Mayuresh wrote:

This may be an OT.


Nah, it's "user" related. Spot on the topic. :-)

I don't think using i386 is a bad idea either, unless one has some 
specific reason to use amd64.


I mostly agree with this sentiment.

I recently switched to amd64. I gained nothing, at least for my purpose, 
but lost wine.


I may be completely misremembering, or perhaps these reasons are merely 
superstitions, but I had a couple of reasons for opting for AMD64. Folks, 
please disabuse me of these notions if they are wrong.


* At some point I thought that PAE wasn't working properly on i386. My
  workstation machine has 24G of RAM (yeah yeah, but it was on sale). I
  have a foggy recollection of i386 only letting me access the first 4GB.

* I get slightly better performance on system benchmarks (again mostly the
  memory benchmarks) with AMD64. Interestingly, lately with NetBSD 7.0
  some benchmarks are beating Linux across the board on the same hardware
  (I use removable drives on the same workstation). I was very pleasantly
  surprised at that.

However,

* It seems like at some point flash was working on i386 but not on AMD64,
  but that may have been a long time ago.

* As you point out wine seems perpetually broken on AMD64. I'm not
  sure why, since it works on Linux x86_64. However, I'm not trying to
  make light of anyone's efforts. Everything takes work.

* Also some time in the past I had major USB bus error problems with AMD64
  but I'd try i386 and the system behaved fine.

All in all, Mayuresh, while I can see some reasons to run AMD64, I'm 
really on the same wavelength as you are - what's the point of hassling 
with AMD64 anyway ? Perhaps someone will point out the "killer feature" 
that we missed. :-)


-Swift

Re: How to run Microsoft Internet Explorer on NetBSD?

[Swift Griggs]

On Fri, 29 Jan 2016, Mayuresh wrote:

On Fri, Jan 29, 2016 at 11:31:17AM +0100, Jose Luis Rodriguez Garcia wrote:

I need to use Internet Explorer for access a web from my company:

www/ies4linux? [ I haven't used of late. ]


That is a useful package, but it won't work if you are on NetBSD 7.0 AMD64 
since Wine is -currently broken for AMD64, IIRC. Last time I checked was 
about a week ago.


Thanks,
  Swift

Re: How to run Microsoft Internet Explorer on NetBSD?

[Mayuresh]

On Fri, Jan 29, 2016 at 09:13:29AM -0700, Swift Griggs wrote:
> On Fri, 29 Jan 2016, Mayuresh wrote:
> >On Fri, Jan 29, 2016 at 11:31:17AM +0100, Jose Luis Rodriguez Garcia wrote:
> >>I need to use Internet Explorer for access a web from my company:
> >www/ies4linux? [ I haven't used of late. ]
> 
> That is a useful package, but it won't work if you are on NetBSD 7.0 AMD64
> since Wine is -currently broken for AMD64, IIRC. Last time I checked was
> about a week ago.

This may be an OT.

I don't think using i386 is a bad idea either, unless one has some
specific reason to use amd64.

I was a long term user of i386 and just to find out whether it makes any
difference (since I could not find any reasons why people use amd64,
except where they know they need a 64 bit system) I recently switched to
amd64. I gained nothing, at least for my purpose, but lost wine.

Mayuresh.

Re: How to run Microsoft Internet Explorer on NetBSD?

[Mayuresh]

On Fri, Jan 29, 2016 at 09:58:52AM -0700, Swift Griggs wrote:
> * I get slightly better performance on system benchmarks (again mostly the
>   memory benchmarks) with AMD64. Interestingly, lately with NetBSD 7.0
>   some benchmarks are beating Linux across the board on the same hardware
>   (I use removable drives on the same workstation). I was very pleasantly
>   surprised at that.

That deserves a great visibility. Could you share the benchmarks?

[ I use both Linux (at work) and NetBSD (at home). I find it difficult to
reason to myself, why don't I simply use Linux everywhere as it is faster,
has greater application availability etc. My instincts always tell me to
continue with NetBSD, but that's not `reasoning' really... ]

Mayuresh.

Re: How to run Microsoft Internet Explorer on NetBSD?

[Hal Murray]

swiftgri...@gmail.com said:
> All in all, Mayuresh, while I can see some reasons to run AMD64, I'm  really
> on the same wavelength as you are - what's the point of hassling  with AMD64
> anyway ? Perhaps someone will point out the "killer feature"  that we
> missed. :-) 

The thing you get with 64 bits is pointers that work in more than 4 gigabytes 
of memory.  That includes virtual memory as well as physical.  There are lots 
of cases where that is critical.  The usual one is databases.

On the other hand, all your pointers take twice as much memory.  That may 
have significant impacts on your cache footprint.

There are some problems where you need 64 bit integers.  You can get those on 
a 32 bit system with long long.  That may be a pain if you are trying to 
compile software that somebody else wrote.


swiftgri...@gmail.com said:
> * At some point I thought that PAE wasn't working properly on i386. My
>workstation machine has 24G of RAM (yeah yeah, but it was on sale). I
>have a foggy recollection of i386 only letting me access the first 4GB. 

I have amd64 7.0 running on a system with 4 GB but /proc/meminfo says it's 
only using 3.  Is there anything I have to do to get it to use the other GB?


-- 
These are my opinions.  I hate spam.

Re: How to run Microsoft Internet Explorer on NetBSD?

[Swift Griggs]

On Fri, 29 Jan 2016, Hal Murray wrote:
The thing you get with 64 bits is pointers that work in more than 4 
gigabytes of memory.


Yes, of course, but with PAE that shouldn't be a problem. Of course, it 
appears that PAE is still considered experimental in NetBSD, since it's 
not enabled by default.


There are some problems where you need 64 bit integers.  You can get 
those on a 32 bit system with long long.


If you are lucky... Without the new-school integer types in in stdint.h, 
those canonical declarations like "long" or even "long long" can be fuzzy.


That may be a pain if you are trying to compile software that somebody 
else wrote.


Agreed. That's why I wish things like uint16_t ,uint32_t, and uint64_t are 
so nice. However, I think they came with C99 or something like that. So, a 
lot of "legacy" code and code written by people who don't know about the 
new types (or don't care) still has older integer types. Programmers 
usually like determinism. I'm not sure why things went so far off the 
rails with integer sizes et al. Probably has to something to do with 
having so many compilers and CPU types in the 80's and 90's.


I have amd64 7.0 running on a system with 4 GB but /proc/meminfo says 
it's only using 3.  Is there anything I have to do to get it to use the 
other GB?


I've noticed the same thing. I always just figured it was the kernel 
reserving buffers or something. I guess if you switched on the 
experimental PAE code, it you might see a lot more of your RAM. Scope out 
this blog:


https://blog.netbsd.org/tnf/entry/pae_support_for_native_i386

I learned a thing or two when I read it just now and it speaks directly to 
this topic.


-Swift