cppcms-users

[Cppcms-users] Strange memory corruption during copy construction of cppdb::result

[<ele...@ex...>]

Hi Artyom,

I seem to have some kind of memory corruption happening during copy
construction of cppdb::result.

This is the scenario:

An asynchronous service (JSON-RPC server) holds a pointer to a synchronous
service to obtain some results from the database. The synchronous service
lives as long as the asynchronous

attach(new Engine(srv, new RPC(srv, new Engine(srv))), "engine",
"/engine{1}", "/engine(/(.*))?", 1);

Engine is the synchronous service, RPC is the asynchronous.

The Engine constructor:

Engine::Engine(cppcms::service& srv, apps::tracker::RPC* p): Base(srv)
{
    attach(p, "rpc", "/rpc{1}", "/rpc", 0);
}

RPC Constructor:

RPC::RPC(cppcms::service& srv, apps::tracker::Base* p):
json_rpc_server(srv), ctr_(p)
{
    using namespace cppcms::rpc;

    attach(ctr_);

    bind("load_move_item_form", json_method(&RPC::load_move_item_form,
this), method_role);
}

the load_move_item_form function does something like this:

ctr_->init();
cppdb::result r = ctr_->get_all_items(aircraft_id, tbl, 0, false);
ctr_->clear()

the problem is the result r contains garbage, random memory.

the get_all_items function works fine when called from a synchronous
application though.

the backend is SQLite.

This is running on Mageia Linux with stock standard cppcms-1.0.3 and
cppdb-0.3.1

If I loop through the result before returning it from the get_all_items
function, the items are there, but after the copy is constructed in the
caller, it contains garbage.

Any ideas?

Re: [Cppcms-users] Strange memory corruption during copy construction of cppdb::result

[Petr J. <ele...@ex...>]

An update:

Well it seems like it's not a memory corruption, but this is the symptoms:

when calling result::next() it hangs after the 1st iteration. It happens
only after the copy-construction.

Petr

Re: [Cppcms-users] Strange memory corruption during copy construction of cppdb::result

[Petr J. <ele...@ex...>]

Ok, wrong again. i've found out whats happening.

The reason the loop was hanging is because inside the loop I was calling
result::get<string>("colname") but colname is NULL in the database.

It still seems like there is something wrong.

Accessing a null column via result::get should either throw() or return
an empty string. Not hang the application.


On 14/05/2013 7:02 PM, Petr Janda wrote:
> An update:
> 
> Well it seems like it's not a memory corruption, but this is the symptoms:
> 
> when calling result::next() it hangs after the 1st iteration. It happens
> only after the copy-construction.
> 
> Petr
> 
> 
> 
> ------------------------------------------------------------------------------
> AlienVault Unified Security Management (USM) platform delivers complete
> security visibility with the essential security capabilities. Easily and
> efficiently configure, manage, and operate all of your security controls
> from a single console and one unified framework. Download a free trial.
> http://p.sf.net/sfu/alienvault_d2d
> 
> 
> 
> _______________________________________________
> Cppcms-users mailing list
> Cpp...@li...
> https://lists.sourceforge.net/lists/listinfo/cppcms-users
> 


-- 
Please use PGP to encrypt your email to ensure our privacy is respected.