ocamlnet-devel — Developer discussion

[Ocamlnet-devel] Netplex issue with VServer PID namespace

[Thomas C. <cal...@gm...>]

Hi,

We are observing an issue when using OCamlnet netplex in combination with
VServer PID namespaces.
We are using Netplex in the multi-process mode.

Here is what we are doing:
  - start our netplex controller
    - use the post_add_hook to enter a new PID namespace
  - use dynamic workload manager to spawn child workers
    - configured with conn_limit=1

  - launch a loop of client connections
    - this spawns a new worker process for each connection

After several successful connections of the loop, clients cannot connect
anymore.
We observe some worker processes in a defunct/zombie state.
The controller and running worker processes seem deadlocked in some
condition.

When we do not use the post_add_hook to enter a new PID namespace, the
problem cannot be triggered anymore.

Do you have any hint on this?

Many thanks.

Thomas

Re: [Ocamlnet-devel] Missing TLS features in OCamlnet 4.x

[Gerd S. <in...@ge...>]

Am Donnerstag, den 07.05.2015, 10:47 +0200 schrieb Thomas Calderon:
> 
> Hi Gerd,
> 
> 
> On Wed, May 6, 2015 at 10:51 PM, Gerd Stolpmann
> <in...@ge...> wrote:
>         Hi Thomas,
>         
>         Am Mittwoch, den 06.05.2015, 16:31 +0200 schrieb Thomas
>         Calderon:
>         > Hi,
>         >
>         >
>         > I have started working towards supporting OCamlnet 4.x in
>         our project.
>         >
>         >
>         > I had to completely rewrite the TLS code, this was expected
>         due to
>         > OCamlnet switching to GnuTLS.
>         >
>         >
>         > I have noticed the following issues with the current TLS
>         code:
>         >   * dh_params is not used in the code therefore DHE-* suites
>         are not
>         > enabled
>         >     - Using the *gnutls_certificate_set_dh_params* solves
>         the issue
>         > (see attachment for example)
>         
>         Well spotted. I have some concerns about the side effect,
>         though. There
>         is no function to copy a certificate.
> 
> 
> I am not following you on this. Why would you want to copy the
> certificate?

Because it is surprising for the caller when create_config modifies its
arguments by setting the dh params for the certificates. And the only
way to avoid is to make a local copy of the certificate and modify that
instead of the argument.

It is possible to work around by saving the function that creates the
certificate. I'll probably do that.

Gerd


> 
>  
>         
>         >   * support for elliptic curve key exchange seems disabled
>         (ECDHE-*
>         > suites), I have not tried an ECC certificate.
>         >   * support for GCM algorithms is not enabled
>         
>         I'm quite sure that I saw ECDHE and GCM in tests. I'll retest
>         tomorrow.
> It might be related to the GnuTLS version mess on my machine (Ubuntu
> has two versions installed).
> 
> Let me know the results of your test.
> 
> 
> 
>         
>         It is possible to de-configure these features in GnuTLS.
>         
>         >   * sample TLS netplex configuration is missing some ";" for
>         each
>         > sections
>         >
>         >
>         > Do you have the same behavior? It might be linked to my
>         GnuTLS
>         > version.
>         > Did you manage to enable higher-end ciphersuites and PFS
>         suites?
>         
>         Gerd
>         
>         >
>         > Thanks for the feedback.
>         >
>         >
>         > Thomas.
>         >
>         >
>         >
>         >
>         >
>         ------------------------------------------------------------------------------
>         > One dashboard for servers and applications across
>         Physical-Virtual-Cloud
>         > Widest out-of-the-box monitoring support with 50+
>         applications
>         > Performance metrics, stats and reports that give you
>         Actionable Insights
>         > Deep dive visibility with transaction tracing using APM
>         Insight.
>         > http://ad.doubleclick.net/ddm/clk/290420510;117567292;y
>         > _______________________________________________
>         > Ocamlnet-devel mailing list
>         > Oca...@li...
>         > https://lists.sourceforge.net/lists/listinfo/ocamlnet-devel
>         
>         --
>         ------------------------------------------------------------
>         Gerd Stolpmann, Darmstadt, Germany    ge...@ge...
>         My OCaml site:          http://www.camlcity.org
>         Contact details:        http://www.camlcity.org/contact.html
>         Company homepage:       http://www.gerd-stolpmann.de
>         ------------------------------------------------------------
>         
> 
> 

-- 
------------------------------------------------------------
Gerd Stolpmann, Darmstadt, Germany    ge...@ge...
My OCaml site:          http://www.camlcity.org
Contact details:        http://www.camlcity.org/contact.html
Company homepage:       http://www.gerd-stolpmann.de
------------------------------------------------------------

Re: [Ocamlnet-devel] Missing TLS features in OCamlnet 4.x

[Gerd S. <in...@ge...>]

Hi,

I could track the problem down. Since GnuTLS-3.2 the library
keeps /dev/urandom all the time open. This conflicts with how netplex
manages file descriptors (namely closing any unknown when forking), and
is also a known security problem because the child processes do not
reinitialize the random number generators.

If you happen to have GnuTLS-3.4 (btw. avoid 3.4.0, there's a bug
causing GNUTLS_E_PUSH_ERROR all the time), and Linux kernel >= 3.17 it
will use the new getrandom() system call instead. This fixes it.

We probably cannot wait for that. In svn there's now a workaround: after
fork(), GnuTLS is forcedly reinitialized. Also, the call of
gnutls_global_init is now delayed until the first use of the library.

I've attached an ssl scan (ssllabs.com) of the https server (with
GnuTLS-3.4.1), and a priority string "SECURE128:+SECURE192". Looks quite
good (that forward secrecy is not supported everywhere has more to do
with the outdated clients).

Gerd



Am Donnerstag, den 07.05.2015, 15:19 +0200 schrieb Thomas Calderon:
> Hi again,
> 
> 
> I have tried using OCamlnet 4.x TLS on a recent setup (Ubuntu 15.04
> amd64 VM) with GnuTLS version being 3.3.8-3ubuntu3.
> I am unable to have a valid TLS exchange using this setup. I have the
> following trace:
> 
> 
> 
>         [Thu May  7 15:07:38 2015] [debug] [3476:0] Uq_engines:
>         cancel_reading socket_multiplex_controller mplex=807 fd=3
>         [Thu May  7 15:07:38 2015] [debug] [3476:0] Uq_engines:
>         input_done socket_multiplex_controller mplex=807 fd=3
>         [Thu May  7 15:07:38 2015] [debug] [3476:0] Uq_engines:
>         tls_multiplex_controller: update
>         [Thu May  7 15:07:38 2015] [debug] [3476:0] Uq_engines:
>         tls_multiplex_controller: config_adapter recv=true
>         [Thu May  7 15:07:38 2015] [debug] [3476:0] Uq_engines:
>         tls_multiplex_controller: cont_handshake (re)start
>         [Thu May  7 15:07:38 2015] [debug] [3476:0] Netsys_tls:
>         Netsys_tls: hello
>         [Thu May  7 15:07:38 2015] [debug] [3476:0] Uq_engines:
>         tls_adapter: recv caller_size=5 avail_size=305 n=5
>         [Thu May  7 15:07:38 2015] [debug] [3476:0] Uq_engines:
>         tls_adapter: recv caller_size=300 avail_size=300 n=300
>         [Thu May  7 15:07:38 2015] [debug] [3476:0] Netsys_tls:
>         Exception in function Netsys_tls.state_driven_action:
>         Nettls_gnutls.TLS.Error(GNUTLS_E_RANDOM_DEVICE_ERROR) -
>         backtrace:
>         [Thu May  7 15:07:38 2015] [debug] [3476:0] Uq_engines:
>         tls_multiplex_controller: cont_handshake
>         exn=Nettls_gnutls.TLS.Error(GNUTLS_E_RANDOM_DEVICE_ERROR)
>         
>         
> Other C programs linked to GnuTLS work fine on this setup.
> Any ideas?
> 
> 
> Cheers.
> 
> On Thu, May 7, 2015 at 10:47 AM, Thomas Calderon
> <cal...@gm...> wrote:
>         
>         Hi Gerd,
>         
>         
>         On Wed, May 6, 2015 at 10:51 PM, Gerd Stolpmann
>         <in...@ge...> wrote:
>                 Hi Thomas,
>                 
>                 Am Mittwoch, den 06.05.2015, 16:31 +0200 schrieb
>                 Thomas Calderon:
>                 > Hi,
>                 >
>                 >
>                 > I have started working towards supporting OCamlnet
>                 4.x in our project.
>                 >
>                 >
>                 > I had to completely rewrite the TLS code, this was
>                 expected due to
>                 > OCamlnet switching to GnuTLS.
>                 >
>                 >
>                 > I have noticed the following issues with the current
>                 TLS code:
>                 >   * dh_params is not used in the code therefore
>                 DHE-* suites are not
>                 > enabled
>                 >     - Using the *gnutls_certificate_set_dh_params*
>                 solves the issue
>                 > (see attachment for example)
>                 
>                 Well spotted. I have some concerns about the side
>                 effect, though. There
>                 is no function to copy a certificate.
>         
>         
>         I am not following you on this. Why would you want to copy the
>         certificate?
>         
>          
>                 
>                 >   * support for elliptic curve key exchange seems
>                 disabled (ECDHE-*
>                 > suites), I have not tried an ECC certificate.
>                 >   * support for GCM algorithms is not enabled
>                 
>                 I'm quite sure that I saw ECDHE and GCM in tests. I'll
>                 retest tomorrow.
>         It might be related to the GnuTLS version mess on my machine
>         (Ubuntu has two versions installed).
>         
>         Let me know the results of your test.
>         
>         
>         
>                 
>                 It is possible to de-configure these features in
>                 GnuTLS.
>                 
>                 >   * sample TLS netplex configuration is missing some
>                 ";" for each
>                 > sections
>                 >
>                 >
>                 > Do you have the same behavior? It might be linked to
>                 my GnuTLS
>                 > version.
>                 > Did you manage to enable higher-end ciphersuites and
>                 PFS suites?
>                 
>                 Gerd
>                 
>                 >
>                 > Thanks for the feedback.
>                 >
>                 >
>                 > Thomas.
>                 >
>                 >
>                 >
>                 >
>                 >
>                 ------------------------------------------------------------------------------
>                 > One dashboard for servers and applications across
>                 Physical-Virtual-Cloud
>                 > Widest out-of-the-box monitoring support with 50+
>                 applications
>                 > Performance metrics, stats and reports that give you
>                 Actionable Insights
>                 > Deep dive visibility with transaction tracing using
>                 APM Insight.
>                 >
>                 http://ad.doubleclick.net/ddm/clk/290420510;117567292;y
>                 > _______________________________________________
>                 > Ocamlnet-devel mailing list
>                 > Oca...@li...
>                 >
>                 https://lists.sourceforge.net/lists/listinfo/ocamlnet-devel
>                 
>                 --
>                 ------------------------------------------------------------
>                 Gerd Stolpmann, Darmstadt, Germany
>                 ge...@ge...
>                 My OCaml site:          http://www.camlcity.org
>                 Contact details:
>                 http://www.camlcity.org/contact.html
>                 Company homepage:       http://www.gerd-stolpmann.de
>                 ------------------------------------------------------------
>                 
>         
>         
> 
> 

-- 
------------------------------------------------------------
Gerd Stolpmann, Darmstadt, Germany    ge...@ge...
My OCaml site:          http://www.camlcity.org
Contact details:        http://www.camlcity.org/contact.html
Company homepage:       http://www.gerd-stolpmann.de
------------------------------------------------------------

Re: [Ocamlnet-devel] Missing TLS features in OCamlnet 4.x

[Thomas C. <cal...@gm...>]

Hi again,

I have tried using OCamlnet 4.x TLS on a recent setup (Ubuntu 15.04 amd64
VM) with GnuTLS version being 3.3.8-3ubuntu3.
I am unable to have a valid TLS exchange using this setup. I have the
following trace:

[Thu May  7 15:07:38 2015] [debug] [3476:0] Uq_engines: cancel_reading
socket_multiplex_controller mplex=807 fd=3
[Thu May  7 15:07:38 2015] [debug] [3476:0] Uq_engines: input_done
socket_multiplex_controller mplex=807 fd=3
[Thu May  7 15:07:38 2015] [debug] [3476:0] Uq_engines:
tls_multiplex_controller: update
[Thu May  7 15:07:38 2015] [debug] [3476:0] Uq_engines:
tls_multiplex_controller: config_adapter recv=true
[Thu May  7 15:07:38 2015] [debug] [3476:0] Uq_engines:
tls_multiplex_controller: cont_handshake (re)start
[Thu May  7 15:07:38 2015] [debug] [3476:0] Netsys_tls: Netsys_tls: hello
[Thu May  7 15:07:38 2015] [debug] [3476:0] Uq_engines: tls_adapter: recv
caller_size=5 avail_size=305 n=5
[Thu May  7 15:07:38 2015] [debug] [3476:0] Uq_engines: tls_adapter: recv
caller_size=300 avail_size=300 n=300
[Thu May  7 15:07:38 2015] [debug] [3476:0] Netsys_tls: Exception in
function Netsys_tls.state_driven_action:
Nettls_gnutls.TLS.Error(GNUTLS_E_RANDOM_DEVICE_ERROR) - backtrace:
[Thu May  7 15:07:38 2015] [debug] [3476:0] Uq_engines:
tls_multiplex_controller: cont_handshake
exn=Nettls_gnutls.TLS.Error(GNUTLS_E_RANDOM_DEVICE_ERROR)

Other C programs linked to GnuTLS work fine on this setup.
Any ideas?

Cheers.

On Thu, May 7, 2015 at 10:47 AM, Thomas Calderon <cal...@gm...>
wrote:

>
> Hi Gerd,
>
>
> On Wed, May 6, 2015 at 10:51 PM, Gerd Stolpmann <in...@ge...>
> wrote:
>
>> Hi Thomas,
>>
>> Am Mittwoch, den 06.05.2015, 16:31 +0200 schrieb Thomas Calderon:
>> > Hi,
>> >
>> >
>> > I have started working towards supporting OCamlnet 4.x in our project.
>> >
>> >
>> > I had to completely rewrite the TLS code, this was expected due to
>> > OCamlnet switching to GnuTLS.
>> >
>> >
>> > I have noticed the following issues with the current TLS code:
>> >   * dh_params is not used in the code therefore DHE-* suites are not
>> > enabled
>> >     - Using the *gnutls_certificate_set_dh_params* solves the issue
>> > (see attachment for example)
>>
>> Well spotted. I have some concerns about the side effect, though. There
>> is no function to copy a certificate.
>>
>
> I am not following you on this. Why would you want to copy the certificate?
>
>
>>
>> >   * support for elliptic curve key exchange seems disabled (ECDHE-*
>> > suites), I have not tried an ECC certificate.
>> >   * support for GCM algorithms is not enabled
>>
>> I'm quite sure that I saw ECDHE and GCM in tests. I'll retest tomorrow.
>>
> It might be related to the GnuTLS version mess on my machine (Ubuntu has
> two versions installed).
> Let me know the results of your test.
>
>
>> It is possible to de-configure these features in GnuTLS.
>>
>> >   * sample TLS netplex configuration is missing some ";" for each
>> > sections
>> >
>> >
>> > Do you have the same behavior? It might be linked to my GnuTLS
>> > version.
>> > Did you manage to enable higher-end ciphersuites and PFS suites?
>>
>> Gerd
>>
>> >
>> > Thanks for the feedback.
>> >
>> >
>> > Thomas.
>> >
>> >
>> >
>> >
>> >
>> ------------------------------------------------------------------------------
>> > One dashboard for servers and applications across Physical-Virtual-Cloud
>> > Widest out-of-the-box monitoring support with 50+ applications
>> > Performance metrics, stats and reports that give you Actionable Insights
>> > Deep dive visibility with transaction tracing using APM Insight.
>> > http://ad.doubleclick.net/ddm/clk/290420510;117567292;y
>> > _______________________________________________
>> > Ocamlnet-devel mailing list
>> > Oca...@li...
>> > https://lists.sourceforge.net/lists/listinfo/ocamlnet-devel
>>
>> --
>> ------------------------------------------------------------
>> Gerd Stolpmann, Darmstadt, Germany    ge...@ge...
>> My OCaml site:          http://www.camlcity.org
>> Contact details:        http://www.camlcity.org/contact.html
>> Company homepage:       http://www.gerd-stolpmann.de
>> ------------------------------------------------------------
>>
>>
>

Re: [Ocamlnet-devel] Missing TLS features in OCamlnet 4.x

[Thomas C. <cal...@gm...>]

Hi Gerd,


On Wed, May 6, 2015 at 10:51 PM, Gerd Stolpmann <in...@ge...>
wrote:

> Hi Thomas,
>
> Am Mittwoch, den 06.05.2015, 16:31 +0200 schrieb Thomas Calderon:
> > Hi,
> >
> >
> > I have started working towards supporting OCamlnet 4.x in our project.
> >
> >
> > I had to completely rewrite the TLS code, this was expected due to
> > OCamlnet switching to GnuTLS.
> >
> >
> > I have noticed the following issues with the current TLS code:
> >   * dh_params is not used in the code therefore DHE-* suites are not
> > enabled
> >     - Using the *gnutls_certificate_set_dh_params* solves the issue
> > (see attachment for example)
>
> Well spotted. I have some concerns about the side effect, though. There
> is no function to copy a certificate.
>

I am not following you on this. Why would you want to copy the certificate?


>
> >   * support for elliptic curve key exchange seems disabled (ECDHE-*
> > suites), I have not tried an ECC certificate.
> >   * support for GCM algorithms is not enabled
>
> I'm quite sure that I saw ECDHE and GCM in tests. I'll retest tomorrow.
>
It might be related to the GnuTLS version mess on my machine (Ubuntu has
two versions installed).
Let me know the results of your test.


> It is possible to de-configure these features in GnuTLS.
>
> >   * sample TLS netplex configuration is missing some ";" for each
> > sections
> >
> >
> > Do you have the same behavior? It might be linked to my GnuTLS
> > version.
> > Did you manage to enable higher-end ciphersuites and PFS suites?
>
> Gerd
>
> >
> > Thanks for the feedback.
> >
> >
> > Thomas.
> >
> >
> >
> >
> >
> ------------------------------------------------------------------------------
> > One dashboard for servers and applications across Physical-Virtual-Cloud
> > Widest out-of-the-box monitoring support with 50+ applications
> > Performance metrics, stats and reports that give you Actionable Insights
> > Deep dive visibility with transaction tracing using APM Insight.
> > http://ad.doubleclick.net/ddm/clk/290420510;117567292;y
> > _______________________________________________
> > Ocamlnet-devel mailing list
> > Oca...@li...
> > https://lists.sourceforge.net/lists/listinfo/ocamlnet-devel
>
> --
> ------------------------------------------------------------
> Gerd Stolpmann, Darmstadt, Germany    ge...@ge...
> My OCaml site:          http://www.camlcity.org
> Contact details:        http://www.camlcity.org/contact.html
> Company homepage:       http://www.gerd-stolpmann.de
> ------------------------------------------------------------
>
>

Re: [Ocamlnet-devel] Missing TLS features in OCamlnet 4.x

[Gerd S. <in...@ge...>]

Hi Thomas,

Am Mittwoch, den 06.05.2015, 16:31 +0200 schrieb Thomas Calderon:
> Hi,
> 
> 
> I have started working towards supporting OCamlnet 4.x in our project.
> 
> 
> I had to completely rewrite the TLS code, this was expected due to
> OCamlnet switching to GnuTLS.
> 
> 
> I have noticed the following issues with the current TLS code:
>   * dh_params is not used in the code therefore DHE-* suites are not
> enabled
>     - Using the *gnutls_certificate_set_dh_params* solves the issue
> (see attachment for example)

Well spotted. I have some concerns about the side effect, though. There
is no function to copy a certificate.

>   * support for elliptic curve key exchange seems disabled (ECDHE-*
> suites), I have not tried an ECC certificate.
>   * support for GCM algorithms is not enabled

I'm quite sure that I saw ECDHE and GCM in tests. I'll retest tomorrow.

It is possible to de-configure these features in GnuTLS.

>   * sample TLS netplex configuration is missing some ";" for each
> sections
> 
> 
> Do you have the same behavior? It might be linked to my GnuTLS
> version.
> Did you manage to enable higher-end ciphersuites and PFS suites?

Gerd

> 
> Thanks for the feedback.
> 
> 
> Thomas.
> 
> 
> 
> 
> ------------------------------------------------------------------------------
> One dashboard for servers and applications across Physical-Virtual-Cloud 
> Widest out-of-the-box monitoring support with 50+ applications
> Performance metrics, stats and reports that give you Actionable Insights
> Deep dive visibility with transaction tracing using APM Insight.
> http://ad.doubleclick.net/ddm/clk/290420510;117567292;y
> _______________________________________________
> Ocamlnet-devel mailing list
> Oca...@li...
> https://lists.sourceforge.net/lists/listinfo/ocamlnet-devel

-- 
------------------------------------------------------------
Gerd Stolpmann, Darmstadt, Germany    ge...@ge...
My OCaml site:          http://www.camlcity.org
Contact details:        http://www.camlcity.org/contact.html
Company homepage:       http://www.gerd-stolpmann.de
------------------------------------------------------------

[Ocamlnet-devel] Missing TLS features in OCamlnet 4.x

[Thomas C. <cal...@gm...>]

Hi,

I have started working towards supporting OCamlnet 4.x in our project.

I had to completely rewrite the TLS code, this was expected due to OCamlnet
switching to GnuTLS.

I have noticed the following issues with the current TLS code:
  * dh_params is not used in the code therefore DHE-* suites are not enabled
    - Using the *gnutls_certificate_set_dh_params* solves the issue (see
attachment for example)

  * support for elliptic curve key exchange seems disabled (ECDHE-*
suites), I have not tried an ECC certificate.
  * support for GCM algorithms is not enabled
  * sample TLS netplex configuration is missing some ";" for each sections

Do you have the same behavior? It might be linked to my GnuTLS version.
Did you manage to enable higher-end ciphersuites and PFS suites?

Thanks for the feedback.

Thomas.

[Ocamlnet-devel] [ANN] ocamlnet-4.0.3

[Gerd S. <in...@ge...>]

Hi,

there is a patch release of ocamlnet:

 - supports now GnuTLS-3.4
 - supports builds of GnuTLS where SRP is disabled (e.g. Fedora).

See http://projects.camlcity.org/projects/ocamlnet.html for details.

Gerd
-- 
------------------------------------------------------------
Gerd Stolpmann, Darmstadt, Germany    ge...@ge...
My OCaml site:          http://www.camlcity.org
Contact details:        http://www.camlcity.org/contact.html
Company homepage:       http://www.gerd-stolpmann.de
------------------------------------------------------------

Re: [Ocamlnet-devel] Issue with Unix socket permission

[Thomas C. <cal...@gm...>]

Hi,

Did this get included in Ocamlnet 4.0 ?

I tried to look for a new option to configure Unix socket permission but
did not find it in the documentation.

Thanks.

Thomas C.

On Mon, Dec 8, 2014 at 3:09 PM, Gerd Stolpmann <in...@ge...>
wrote:

> Thanks for reporting this, I've added this to the list of things for
> Ocamlnet-4, which is now likely to be released in the next weeks (eoy
> perhaps).
>
> Gerd
>
> Am Mittwoch, den 19.11.2014, 16:49 +0100 schrieb Thomas Calderon:
> > Hi,
> >
> >
> > I would like to report an issue when using Netplex with a UNIX socket.
> >
> >
> > There is, as far as I know, no way to indicate the socket permission.
> > Therefore, when launching a netplex server process, the socket ends up
> > with default system permission (generally 755).
> > Hence, clients with different UID cannot properly exchange with the
> > server.
> >
> >
> > This is a serious limitation when using this socket backend.
> > It would be preferable to have fine-grained access to the desired
> > socket permissions for instance from the netplex configuration file.
> >
> >
> > Regards,
> >
> >
> > Thomas C.
> >
> ------------------------------------------------------------------------------
> > Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
> > from Actuate! Instantly Supercharge Your Business Reports and Dashboards
> > with Interactivity, Sharing, Native Excel Exports, App Integration & more
> > Get technology previously reserved for billion-dollar corporations, FREE
> >
> http://pubads.g.doubleclick.net/gampad/clk?id=157005751&iu=/4140/ostg.clktrk
> > _______________________________________________
> > Ocamlnet-devel mailing list
> > Oca...@li...
> > https://lists.sourceforge.net/lists/listinfo/ocamlnet-devel
>
> --
> ------------------------------------------------------------
> Gerd Stolpmann, Darmstadt, Germany    ge...@ge...
> My OCaml site:          http://www.camlcity.org
> Contact details:        http://www.camlcity.org/contact.html
> Company homepage:       http://www.gerd-stolpmann.de
> ------------------------------------------------------------
>
>

[Ocamlnet-devel] [ANN] ocamlnet-4.0.2

[Gerd S. <in...@ge...>]

Hi,

just released ocamlnet-4.0.2, with some bug fixes (one important).

Note that I had to change the type of Uq_engines.engine (new method
request_proxy_notification). If you define your own engines, this may
break your builds. Most users won't be affected by this, though.

Download, changelog and other links under
http://projects.camlcity.org/projects/ocamlnet.html

Gerd
-- 
------------------------------------------------------------
Gerd Stolpmann, Darmstadt, Germany    ge...@ge...
My OCaml site:          http://www.camlcity.org
Contact details:        http://www.camlcity.org/contact.html
Company homepage:       http://www.gerd-stolpmann.de
------------------------------------------------------------

Re: [Ocamlnet-devel] Content-Range in Nethttp.Header missing "bytes"

[Gerd S. <in...@ge...>]

Thanks. This is fixed in svn now.

Gerd

Am Montag, den 26.01.2015, 15:25 +0200 schrieb Török Edwin:
> Hi,
> 
> set_content_range in nethttp.ml sets header of the form "Content-Range: 598-3135/3136", but it should set "Content-Range: bytes 598-3135/3136",
> see http://tools.ietf.org/html/rfc7233#section-4.2
> 
> Interestingly the headers set by Nethttp.set_content_range worked fine with clients like curl/wget or even browsers.
> Then I tried using my server as an APT http repository and it complained:
> HTTP/1.1 206 Partial Content
> Accept-Ranges: bytes
> Content-Length: 2538
> Content-Range: 598-3135/3136
> Content-Type: binary/octet-stream
> Date: Sun, 25 Jan 2015 09:51:26 GMT
> ETag: "2015-01-23 10:19:24.611:0ad6cf8d3cfc5ebfc5983d6e8c626bc8"
> Last-Modified: Fri, 23 Jan 2015 10:19:24 GMT
> Server: libres3-0.9.beta-68-g137d4f1-dirty
> x-amz-id-2: dm9sLXJlcG8AL2RlYmlhbi9kaXN0cy93aGVlenkvSW5SZWxlYXNlADE0MjIxNzk0ODYuNDE=
> x-amz-request-id: libres3_d598fe779a57c81f28f25e1dd9802276_14711_1209_1422179483.37
> 
> 
> 400 URI Failure
> URI: [...]
> Message: The HTTP server sent an invalid Content-Range header
> 
> 
> Testcase (make sure Nethttp can parse the header it generates):
> 
> $ cat >x.ml <<EOF
> #use "topfind"
> #require "netstring"
> let test_bytes b =
>   let  h = new Netmime.basic_mime_header [] in
>   Nethttp.Header.set_content_range h b;
>   let cr = h#field "Content-Range" in
> (*  if String.length cr < 6 || String.sub cr 0 6 <> "bytes" then
>     h#update_field "Content-Range" ("bytes " ^ cr);
>   let cr = h#field "Content-Range" in*)
>   prerr_endline cr;
>   assert (Nethttp.Header.get_content_range h = b)
> 
> let () =
>   test_bytes (`Bytes ((Some (42L, 1233L), Some 1234L)));
>   test_bytes (`Bytes ((Some (42L, 1233L), None)));
>   test_bytes (`Bytes (None, Some 1234L));
> EOF
> $ ocaml x.ml
> 42-1233/1234
> Exception: Nethttp.Bad_header_field "Nethttp.get_content_range".
> 
> Best regards,
> --Edwin
> 
> ------------------------------------------------------------------------------
> Dive into the World of Parallel Programming. The Go Parallel Website,
> sponsored by Intel and developed in partnership with Slashdot Media, is your
> hub for all things parallel software development, from weekly thought
> leadership blogs to news, videos, case studies, tutorials and more. Take a
> look and join the conversation now. http://goparallel.sourceforge.net/
> _______________________________________________
> Ocamlnet-devel mailing list
> Oca...@li...
> https://lists.sourceforge.net/lists/listinfo/ocamlnet-devel

-- 
------------------------------------------------------------
Gerd Stolpmann, Darmstadt, Germany    ge...@ge...
My OCaml site:          http://www.camlcity.org
Contact details:        http://www.camlcity.org/contact.html
Company homepage:       http://www.gerd-stolpmann.de
------------------------------------------------------------

[Ocamlnet-devel] Content-Range in Nethttp.Header missing "bytes"

[Török E. <edw...@et...>]

Hi,

set_content_range in nethttp.ml sets header of the form "Content-Range: 598-3135/3136", but it should set "Content-Range: bytes 598-3135/3136",
see http://tools.ietf.org/html/rfc7233#section-4.2

Interestingly the headers set by Nethttp.set_content_range worked fine with clients like curl/wget or even browsers.
Then I tried using my server as an APT http repository and it complained:
HTTP/1.1 206 Partial Content
Accept-Ranges: bytes
Content-Length: 2538
Content-Range: 598-3135/3136
Content-Type: binary/octet-stream
Date: Sun, 25 Jan 2015 09:51:26 GMT
ETag: "2015-01-23 10:19:24.611:0ad6cf8d3cfc5ebfc5983d6e8c626bc8"
Last-Modified: Fri, 23 Jan 2015 10:19:24 GMT
Server: libres3-0.9.beta-68-g137d4f1-dirty
x-amz-id-2: dm9sLXJlcG8AL2RlYmlhbi9kaXN0cy93aGVlenkvSW5SZWxlYXNlADE0MjIxNzk0ODYuNDE=
x-amz-request-id: libres3_d598fe779a57c81f28f25e1dd9802276_14711_1209_1422179483.37


400 URI Failure
URI: [...]
Message: The HTTP server sent an invalid Content-Range header


Testcase (make sure Nethttp can parse the header it generates):

$ cat >x.ml <<EOF
#use "topfind"
#require "netstring"
let test_bytes b =
  let  h = new Netmime.basic_mime_header [] in
  Nethttp.Header.set_content_range h b;
  let cr = h#field "Content-Range" in
(*  if String.length cr < 6 || String.sub cr 0 6 <> "bytes" then
    h#update_field "Content-Range" ("bytes " ^ cr);
  let cr = h#field "Content-Range" in*)
  prerr_endline cr;
  assert (Nethttp.Header.get_content_range h = b)

let () =
  test_bytes (`Bytes ((Some (42L, 1233L), Some 1234L)));
  test_bytes (`Bytes ((Some (42L, 1233L), None)));
  test_bytes (`Bytes (None, Some 1234L));
EOF
$ ocaml x.ml
42-1233/1234
Exception: Nethttp.Bad_header_field "Nethttp.get_content_range".

Best regards,
--Edwin

[Ocamlnet-devel] [ANN] new releases: pxp, webdav, cache, netamqp, wdialog

[Gerd S. <in...@ge...>]

Hi list,

as followup to OCamlnet-4.0.1 there are also new releases of these
libraries that have been ported to OCamlnet-4.0.1:

http://projects.camlcity.org/projects/pxp.html
http://projects.camlcity.org/projects/webdav.html
http://projects.camlcity.org/projects/cache.html
http://projects.camlcity.org/projects/netamqp.html
http://projects.camlcity.org/projects/wdialog.html

Gerd
-- 
------------------------------------------------------------
Gerd Stolpmann, Darmstadt, Germany    ge...@ge...
My OCaml site:          http://www.camlcity.org
Contact details:        http://www.camlcity.org/contact.html
Company homepage:       http://www.gerd-stolpmann.de
------------------------------------------------------------

[Ocamlnet-devel] [ANN] ocamlnet-4.0.1

[Gerd S. <in...@ge...>]

Hi list,

after some testing I could finally release OCamlnet-4.0.1. Find it at
the usual place:

http://projects.camlcity.org/projects/ocamlnet.html

This is a major release, and there are incompatible changes, in
particular in the following areas:

 - How ocamlnet is configured (new switches: -enable-gnutls,
   -enable-gssapi, gone: -enable-ssl, -enable-crypto)
 - How to use TLS (formerly SSL) in protocols (see below)
 - Unicode tables need to be explicitly linked in (package
   netunidata), or conversions between character encodings
   may be unavailable
 - IPv6 is now enabled by default if there is a global IPv6
   address
 - A number of modules has been renamed (e.g. Http_client is
   now Nethttp_client)
 - A number of modules have been split up (e.g. some functions
   of Uq_engines are now in Uq_client and other modules)

But these are only the most visible changes. Besides that, there are
also plenty of new features, all about authentication and security.

1. TLS

The old support for OpenSSL (via the ocaml ssl bindings) are no longer
supported. Instead, there is an integrated wrapper for GnuTLS, which has
much nicer API and has better support for the latest TLS updates.

Read more here:
http://projects.camlcity.org/projects/dl/ocamlnet-4.0.1/doc/html-main/Tls.html

2. SASL authentication

OCamlnet-4 implements all important SASL authentication mechanisms (for
both clients and servers): PLAIN, CRAM-MD5, DIGEST-MD5, SCRAM-SHA1,
GS2-KRB5 (more on that one below). SASL is integrated into the protocol
interpreters for POP and  SMTP.

3. GSSAPI/Kerberos

Before this release, the was only very preliminary support for the
GSSAPI. This has changed, as there is now a binding of the system-level
GSSAPI functions. In particular, this permits protocol interpreters to
support Kerberos, and in Ocamlnet-4 SunRPC, HTTP clients, and FTP have
been "kerberized" this way. You can also run Kerberos on top of SASL via
bridging.

If you are unfamiliar with Kerberos, this is a single-sign-on security
framework for LANs, and it can be frequently found in company networks.

At the moment, we cannot support Kerberos on Windows, though (as Windows
doesn't provide a standard-compliant GSSAPI library).

4. Pluggable security modules

The security mechanisms are all pluggable. If you e.g. want to use a
different TLS provider, you "only" need to ensure that it has a module
interface compatible with TLS_PROVIDER. There's also SASL_MECHANISM,
HTTP_MECHANISM, and GSSAPI. The mechanisms are passed as normal values
to the protocol interpreters thanks to first-class modules. This permits
it to support several implementations, and also to weakly bind to
security mechanisms (e.g. use only TLS if available).

Gerd
-- 
------------------------------------------------------------
Gerd Stolpmann, Darmstadt, Germany    ge...@ge...
My OCaml site:          http://www.camlcity.org
Contact details:        http://www.camlcity.org/contact.html
Company homepage:       http://www.gerd-stolpmann.de
------------------------------------------------------------

Re: [Ocamlnet-devel] looking for a minimalistic example of RPC

[Gerd S. <in...@ge...>]

Am Montag, den 12.01.2015, 17:49 +0100 schrieb Francois Berenger:
> Hello,
> 
> Where can I find a minimalistic example of a client and server
> using ocamlnet's RPC library?
> 
> I looked into the sources of ocamlnet in examples/rpc/finder
> but the finder example uses netplex, so it is not really
> a minimalistic example.
> 
> I am looking for something really trivial, for example:
> 
> A module A with function 'some_int ()' that returns
> an int each time it is called.
> A module B, client of A which would invoke the remote some_int
> method provided by module A and get the result in a synchronous manner.

The "queues" examples is of that type (well, it's not using netplex, and
the clients are synchronous, and there are also synchronous server
functions, e.g. create_queue). The data types are a bit more complicated
than just ints, though.

You can also look into tests/rpc/bench, it's even simpler. (I think
tests/ is only in the svn repo.)

Gerd
-- 
------------------------------------------------------------
Gerd Stolpmann, Darmstadt, Germany    ge...@ge...
My OCaml site:          http://www.camlcity.org
Contact details:        http://www.camlcity.org/contact.html
Company homepage:       http://www.gerd-stolpmann.de
------------------------------------------------------------

[Ocamlnet-devel] looking for a minimalistic example of RPC

[Francois B. <fra...@in...>]

Hello,

Where can I find a minimalistic example of a client and server
using ocamlnet's RPC library?

I looked into the sources of ocamlnet in examples/rpc/finder
but the finder example uses netplex, so it is not really
a minimalistic example.

I am looking for something really trivial, for example:

A module A with function 'some_int ()' that returns
an int each time it is called.
A module B, client of A which would invoke the remote some_int
method provided by module A and get the result in a synchronous manner.

-- 
Regards,
Francois.

[Ocamlnet-devel] [ANN] ocamlnet-4.0.0rc1

[Gerd S. <in...@ge...>]

Hi list,

with the first release candidate the OCamlnet development cycle goes
into the final round. If no serious problems are found, this version
will be released (in about two weeks). If you'd like to help testing
this version, this is now the best time, in particular for checking
whether the build is ok on all platforms. OCamlnet contains now bindings
for two new libraries: GnuTLS and GSSAPI, and these new bindings could
cause problems here and there.

ocamlnet-4.0.0rc1 is available at the usual place:
http://projects.camlcity.org/projects/ocamlnet.html

When you install it, be prepared that the configure script takes
somewhat different arguments than before (new: -enable-gnutls,
-enable-gssapi, gone: -enable-ssl, -enable-crypto).

So, what's new? OCamlnet-4 focuses on authentication and security. There
is now excellent support for TLS via the new GnuTLS bindings. It's
available for all protocols, it's recent, and it supports the newest TLS
options (e.g. TLS-1.2 with ECCDH). You can query the TLS connection
properties. You can check the certificates (there is a parser for
certificates). Via GnuTLS we also get direct access to hash functions
and symmetric ciphers (sometimes even accelerated).

There is an option for client certificates, but most protocols prefer to
log users in with passwords. OCamlnet-4 implements all important SASL
authentication mechanisms (for both clients and servers): PLAIN,
CRAM-MD5, DIGEST-MD5, SCRAM-SHA1, GS2-KRB5 (more on that one below).
SASL is integrated into the protocol interpreters for POP and SMTP, but
it's quite simple to use, so feel free to include it into your own
protocols. There is support for the SASLprep routine that prepares and
normalizes passwords.

The authentication mechanisms for the HTTP client have been updated, but
this overlaps with the current IETF activities (Digest authentication is
being revised, and SCRAM authentication being defined by IETF working
groups). When the new mechanisms are available, OCamlnet can implement
them very quickly.

The GSSAPI bindings allows us to access network authentication
mechanisms, in particular Kerberos which is mainly found in LANs of
companies and organizations. The GSSAPI is integrated into all protocols
where it is defined for, in particular SunRPC, HTTP (client) i.e.
SPNEGO, and FTP. The GSSAPI mechanisms can also be used as SASL
mechanisms (via bridging, GS2-KRB5). With Kerberos, the most important
GSSAPI mechanism, you get automatic logins to network services (no
password required, or better Kerberos knows how to use the current
system identity in the network - single-sign-on).

The security mechanisms are all pluggable. If you e.g. want to use a
different TLS provider, you "only" need to ensure that it has a module
interface compatible with TLS_PROVIDER, and you can use it with
OCamlnet, thanks to first-class modules. There's also SASL_MECHANISM,
HTTP_MECHANISM, and GSSAPI. 

Besides security, a number of other updates are available. In
particular, the IPv6 support is now complete, and IPv6 is automatically
enabled when a network interface has a global IPv6 address. The Unicode
tables have been moved to a library of their own: netunidata. (You may
need to add netunidata to your findlib packages when building your
software against OCamlnet.) A number of modules were renamed, e.g.
Http_client is now Nethttp_client.

Note that 4.0.0rc1 doesn't support the new "bytes" type specially. I've
delayed this because this the current "bytes" type seems not yet not to
be a stable language feature.

Gerd
-- 
------------------------------------------------------------
Gerd Stolpmann, Darmstadt, Germany    ge...@ge...
My OCaml site:          http://www.camlcity.org
Contact details:        http://www.camlcity.org/contact.html
Company homepage:       http://www.gerd-stolpmann.de
------------------------------------------------------------

Re: [Ocamlnet-devel] Issue with Unix socket permission

[Gerd S. <in...@ge...>]

Thanks for reporting this, I've added this to the list of things for
Ocamlnet-4, which is now likely to be released in the next weeks (eoy
perhaps).

Gerd

Am Mittwoch, den 19.11.2014, 16:49 +0100 schrieb Thomas Calderon:
> Hi,
> 
> 
> I would like to report an issue when using Netplex with a UNIX socket.
> 
> 
> There is, as far as I know, no way to indicate the socket permission.
> Therefore, when launching a netplex server process, the socket ends up
> with default system permission (generally 755).
> Hence, clients with different UID cannot properly exchange with the
> server.
> 
> 
> This is a serious limitation when using this socket backend.
> It would be preferable to have fine-grained access to the desired
> socket permissions for instance from the netplex configuration file.
> 
> 
> Regards,
> 
> 
> Thomas C.
> ------------------------------------------------------------------------------
> Download BIRT iHub F-Type - The Free Enterprise-Grade BIRT Server
> from Actuate! Instantly Supercharge Your Business Reports and Dashboards
> with Interactivity, Sharing, Native Excel Exports, App Integration & more
> Get technology previously reserved for billion-dollar corporations, FREE
> http://pubads.g.doubleclick.net/gampad/clk?id=157005751&iu=/4140/ostg.clktrk
> _______________________________________________
> Ocamlnet-devel mailing list
> Oca...@li...
> https://lists.sourceforge.net/lists/listinfo/ocamlnet-devel

-- 
------------------------------------------------------------
Gerd Stolpmann, Darmstadt, Germany    ge...@ge...
My OCaml site:          http://www.camlcity.org
Contact details:        http://www.camlcity.org/contact.html
Company homepage:       http://www.gerd-stolpmann.de
------------------------------------------------------------

[Ocamlnet-devel] [PATCH] Fix Netplex "group" parameter

[Thomas C. <cal...@gm...>]

Hi,

I would like to report an issue regarding the group parameter in a netplex
process.
It can be used to set a group for the worker processes.

It seems the code wrongly re-use the "user" variable instead of using the
"group" variable to fetch the corresponding group with Unix.getgrnam.

This issue is hidden if you have matching user/group (user1/user1), but if
you start using a different user/group pair (user1/nogroup) then
netplex_config fails.

I have attached a patch (applied to 3.7.3) to correct this behavior. It
should be easily adapted to other versions.

Regards,

Thomas C.

[Ocamlnet-devel] Issue with Unix socket permission

[Thomas C. <cal...@gm...>]

Hi,

I would like to report an issue when using Netplex with a UNIX socket.

There is, as far as I know, no way to indicate the socket permission.
Therefore, when launching a netplex server process, the socket ends up with
default system permission (generally 755).
Hence, clients with different UID cannot properly exchange with the server.

This is a serious limitation when using this socket backend.
It would be preferable to have fine-grained access to the desired socket
permissions for instance from the netplex configuration file.

Regards,

Thomas C.

[Ocamlnet-devel] [ANN] ocamlnet-3.7.7

[Gerd S. <in...@ge...>]

Hi,

I just released ocamlnet-3.7.7, with the pending fix for Mac OS X 10.10.
This is the only update.

http://projects.camlcity.org/projects/ocamlnet.html

Gerd
-- 
------------------------------------------------------------
Gerd Stolpmann, Darmstadt, Germany    ge...@ge...
My OCaml site:          http://www.camlcity.org
Contact details:        http://www.camlcity.org/contact.html
Company homepage:       http://www.gerd-stolpmann.de
------------------------------------------------------------

Re: [Ocamlnet-devel] compile error on Mac OS X 10.10 (Yosemite)

[Ashish A. <aga...@gm...>]

I confirm that the trunk version compiles.

On Thu, Oct 23, 2014 at 6:41 AM, Gerd Stolpmann <in...@ge...>
wrote:

> Yes, it is already fixed in svn:
>
> svn diff -r 2088:2089
> https://godirepo.camlcity.org/svn/lib-ocamlnet2/trunk
>
> 10.10 does not define mkfifoat, but the other *at functions, confusing
> previous releases of ocamlnet. I haven't released the patch yet because
> I did not get feedback whether the build really works now. So, if
> somebody could just checkout and try it on 10.10, this would be highly
> appreciated.
>
> Gerd
>
> Am Dienstag, den 21.10.2014, 07:54 -0400 schrieb Ashish Agarwal:
> > Has anyone else hit this compile error. It happens on the latest Mac
> > OS X 10.10. I tested with OCaml 4.02.0 and 4.02.1. I'm compiling via
> > opam, which provides ocamlnet 3.7.6.
> >
> >
> >
> >
> > # ocamlfind ocamlc -g  -I ../../src/netstring -package "unix"  -w
> > -3-25 -c  main.ml
> > # rm -f ocamlrpcgen ocamlrpcgen
> > # OCAMLPATH="..:$OCAMLPATH" \
> > # ocamlfind ocamlc -g  -I ../../src/netstring -package "unix"  -w
> > -3-25 -custom -o ocamlrpcgen \
> > # -package unix,netstring,str \
> > # rpc_generator.cma main.cmo -linkpkg
> > # rm lexer.ml
> > ### stderr ###
> > # ...[truncated]
> > # Undefined symbols for architecture x86_64:
> > #   "_mkfifoat", referenced from:
> > #       _netsys_mkfifoat in libnetsys.a(netsys_c.o)
> > #      (maybe you meant: _netsys_mkfifoat)
> > # ld: symbol(s) not found for architecture x86_64
> > # clang: error: linker command failed with exit code 1 (use -v to see
> > invocation)
> > # File "_none_", line 1:
> > # Error: Error while building custom runtime system
> > # make[1]: *** [ocamlrpcgen] Error 2
> > # make: *** [all] Error 2
> >
> >
> >
> ------------------------------------------------------------------------------
> > Comprehensive Server Monitoring with Site24x7.
> > Monitor 10 servers for $9/Month.
> > Get alerted through email, SMS, voice calls or mobile push notifications.
> > Take corrective actions from your mobile device.
> > http://p.sf.net/sfu/Zoho
> > _______________________________________________
> > Ocamlnet-devel mailing list
> > Oca...@li...
> > https://lists.sourceforge.net/lists/listinfo/ocamlnet-devel
>
> --
> ------------------------------------------------------------
> Gerd Stolpmann, Darmstadt, Germany    ge...@ge...
> My OCaml site:          http://www.camlcity.org
> Contact details:        http://www.camlcity.org/contact.html
> Company homepage:       http://www.gerd-stolpmann.de
> ------------------------------------------------------------
>
>

Re: [Ocamlnet-devel] compile error on Mac OS X 10.10 (Yosemite)

[Gerd S. <in...@ge...>]

Yes, it is already fixed in svn:

svn diff -r 2088:2089
https://godirepo.camlcity.org/svn/lib-ocamlnet2/trunk

10.10 does not define mkfifoat, but the other *at functions, confusing
previous releases of ocamlnet. I haven't released the patch yet because
I did not get feedback whether the build really works now. So, if
somebody could just checkout and try it on 10.10, this would be highly
appreciated.

Gerd

Am Dienstag, den 21.10.2014, 07:54 -0400 schrieb Ashish Agarwal:
> Has anyone else hit this compile error. It happens on the latest Mac
> OS X 10.10. I tested with OCaml 4.02.0 and 4.02.1. I'm compiling via
> opam, which provides ocamlnet 3.7.6.
> 
> 
> 
> 
> # ocamlfind ocamlc -g  -I ../../src/netstring -package "unix"  -w
> -3-25 -c  main.ml
> # rm -f ocamlrpcgen ocamlrpcgen
> # OCAMLPATH="..:$OCAMLPATH" \
> # ocamlfind ocamlc -g  -I ../../src/netstring -package "unix"  -w
> -3-25 -custom -o ocamlrpcgen \
> # -package unix,netstring,str \
> # rpc_generator.cma main.cmo -linkpkg
> # rm lexer.ml
> ### stderr ###
> # ...[truncated]
> # Undefined symbols for architecture x86_64:
> #   "_mkfifoat", referenced from:
> #       _netsys_mkfifoat in libnetsys.a(netsys_c.o)
> #      (maybe you meant: _netsys_mkfifoat)
> # ld: symbol(s) not found for architecture x86_64
> # clang: error: linker command failed with exit code 1 (use -v to see
> invocation)
> # File "_none_", line 1:
> # Error: Error while building custom runtime system
> # make[1]: *** [ocamlrpcgen] Error 2
> # make: *** [all] Error 2
> 
> 
> ------------------------------------------------------------------------------
> Comprehensive Server Monitoring with Site24x7.
> Monitor 10 servers for $9/Month.
> Get alerted through email, SMS, voice calls or mobile push notifications.
> Take corrective actions from your mobile device.
> http://p.sf.net/sfu/Zoho
> _______________________________________________
> Ocamlnet-devel mailing list
> Oca...@li...
> https://lists.sourceforge.net/lists/listinfo/ocamlnet-devel

-- 
------------------------------------------------------------
Gerd Stolpmann, Darmstadt, Germany    ge...@ge...
My OCaml site:          http://www.camlcity.org
Contact details:        http://www.camlcity.org/contact.html
Company homepage:       http://www.gerd-stolpmann.de
------------------------------------------------------------

[Ocamlnet-devel] compile error on Mac OS X 10.10 (Yosemite)

[Ashish A. <aga...@gm...>]

Has anyone else hit this compile error. It happens on the latest Mac OS X
10.10. I tested with OCaml 4.02.0 and 4.02.1. I'm compiling via opam, which
provides ocamlnet 3.7.6.


# ocamlfind ocamlc -g  -I ../../src/netstring -package "unix"  -w -3-25 -c
main.ml
# rm -f ocamlrpcgen ocamlrpcgen
# OCAMLPATH="..:$OCAMLPATH" \
# ocamlfind ocamlc -g  -I ../../src/netstring -package "unix"  -w -3-25
-custom -o ocamlrpcgen \
# -package unix,netstring,str \
# rpc_generator.cma main.cmo -linkpkg
# rm lexer.ml
### stderr ###
# ...[truncated]
# Undefined symbols for architecture x86_64:
#   "_mkfifoat", referenced from:
#       _netsys_mkfifoat in libnetsys.a(netsys_c.o)
#      (maybe you meant: _netsys_mkfifoat)
# ld: symbol(s) not found for architecture x86_64
# clang: error: linker command failed with exit code 1 (use -v to see
invocation)
# File "_none_", line 1:
# Error: Error while building custom runtime system
# make[1]: *** [ocamlrpcgen] Error 2
# make: *** [all] Error 2

[Ocamlnet-devel] [ANN] ocamlnet-3.7.6

[Gerd S. <in...@ge...>]

Hello,

I just released ocamlnet-3.7.6. It fixes only one problem, namely that
netstring-pcre can be built without camlp4. In 3.7.5 camplp4 was still
required although the docs promised something different.

http://projects.camlcity.org/projects/ocamlnet.html

Gerd
-- 
------------------------------------------------------------
Gerd Stolpmann, Darmstadt, Germany    ge...@ge...
My OCaml site:          http://www.camlcity.org
Contact details:        http://www.camlcity.org/contact.html
Company homepage:       http://www.gerd-stolpmann.de
------------------------------------------------------------