Re: [lwip-devel] [bug #50040] pbuf_alloc(..., 65534, PBUF_RAM) succedes

[D.C. van Moolenbroek]

On 1/12/2017 20:28, address@hidden wrote:
> David van Moolenbroek wrote:
> > [Off-topic: that udp case does worry me: even if allocation doesn't get
> > truncated, it will hit the assertion I mentioned, but lwIP doesn't 
> > seem to
> > have a clear model for "use PBUF_POOL for allocation if we are using 
> > pools, or
> > PBUF_RAM otherwise" so I am not sure what should be done there. There 
> > are
> > probably other cases like that, and this does worry me quite a bit..]
>
> Up to now, PBUF_POOL is actually meant for RX from a netif. All other 
> cases use PBUF_RAM for allocation.
>
> This is to be improved (see the unfortunately rather old buf #3031).
> The idea is to convert 'pbuf_type' to a bit field e.g. indicating 
> tx/rx, contiguous buffer requested, etc.
> and then letting the code behind (or even the application) decide 
> which memory type or which pool is used... 

Well, that sounds great, but it also sounds like a massive change where 
it's going to be very hard to keep into account everyone's situation etc 
- which may explain why it hasn't happened yet.. I really can't promise 
being able to work on anything myself, but it seems that it would be 
possible to make at least some progress in the right direction, for 
example by adding exactly such a "POOL chain or RAM depending on your 
configuration" allocation option and preparing a bit more code for the 
possibility of getting back a chain of pbufs rather than a single one. 
Do such more incremental changes still sound like a welcome step forward?

David