lwip-users
[Top][All Lists]
Advanced
[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

[lwip-users] [lwip] free-ing of pbufs in netif->output function: who's

From: Lieverse, Paul
Subject: [lwip-users] [lwip] free-ing of pbufs in netif->output function: who's responsible?
Date: Thu, 09 Jan 2003 00:25:21 -0000 
This message is in MIME format. Since your mail reader does not understand
this format, some or all of this message may not be legible.

--------------InterScan_NT_MIME_Boundary
Content-Type: multipart/alternative;
        boundary="----_=_NextPart_001_01C1C11F.BD106600"

------_=_NextPart_001_01C1C11F.BD106600
Content-Type: text/plain

Hi,

I was wondering how the pbuf management in the netif drivers should work.
To be more precise: should the drivers do a pbuf_free on the pbufs that they
get passed, or only on the pbufs they allocate themselves? (Currently,
neither
is done, which is, as far as I can see, wrong anyway).

>From the code I understand that:
1) TCP/UDP code passes a pbuf pointer to ip_output
2) ip_ouput passes this pointer to the netif output function
3) eventually, it will be passed to some low_level_output function.

Now who is responsible for a pbuf that has been sent out?

Some observations:
*) tcp_output() and tcp_rst() do a pbuf_free after ip_output()
*) tcp_output_segment() and tcp_rexmit_seg() don't do a pbuf_free
*) udp_send() doesn't do a pbuf_free either

And to make things more complicated:
*) In several functions (e.g. in the netif output functions) some extra pbuf
   is allocated and chained at the head of the given pbuf. Noone
   seems to be free-ing those.

So, am I correct that some memory may be leaking here, or am I really
missing something?

Paul

------_=_NextPart_001_01C1C11F.BD106600
Content-Type: text/html

<!DOCTYPE HTML PUBLIC "-//W3C//DTD HTML 3.2//EN">
<HTML>
<HEAD>
<META HTTP-EQUIV="Content-Type" CONTENT="text/html; charset=US-ASCII">
<META NAME="Generator" CONTENT="MS Exchange Server version 5.5.2653.12">
<TITLE>free-ing of pbufs in netif-&gt;output function: who's 
responsible?</TITLE>
</HEAD>
<BODY>

<P><FONT SIZE=2 FACE="Arial">Hi,</FONT>
</P>

<P><FONT SIZE=2 FACE="Arial">I was wondering how the pbuf management in the 
netif drivers should work.</FONT>
<BR><FONT SIZE=2 FACE="Arial">To be more precise: should the drivers do a 
pbuf_free on the pbufs that they</FONT>
<BR><FONT SIZE=2 FACE="Arial">get passed, or only on the pbufs they allocate 
themselves? (Currently, neither</FONT>
<BR><FONT SIZE=2 FACE="Arial">is done, which is, as far as I can see, wrong 
anyway).</FONT>
</P>

<P><FONT SIZE=2 FACE="Arial">From the code I understand that:</FONT>
<BR><FONT SIZE=2 FACE="Arial">1) TCP/UDP code passes a pbuf pointer to 
ip_output</FONT>
<BR><FONT SIZE=2 FACE="Arial">2) ip_ouput passes this pointer to the netif 
output function</FONT>
<BR><FONT SIZE=2 FACE="Arial">3) eventually, it will be passed to some 
low_level_output function.</FONT>
</P>

<P><FONT SIZE=2 FACE="Arial">Now who is responsible for a pbuf that has been 
sent out?</FONT>
</P>

<P><FONT SIZE=2 FACE="Arial">Some observations:</FONT>
<BR><FONT SIZE=2 FACE="Arial">*) tcp_output() and tcp_rst() do a pbuf_free 
after ip_output()</FONT>
<BR><FONT SIZE=2 FACE="Arial">*) tcp_output_segment() and tcp_rexmit_seg() 
don't do a pbuf_free</FONT>
<BR><FONT SIZE=2 FACE="Arial">*) udp_send() doesn't do a pbuf_free either</FONT>
</P>

<P><FONT SIZE=2 FACE="Arial">And to make things more complicated:</FONT>
<BR><FONT SIZE=2 FACE="Arial">*) In several functions (e.g. in the netif output 
functions) some extra pbuf</FONT>
<BR><FONT SIZE=2 FACE="Arial">&nbsp;&nbsp; is allocated and chained at the head 
of the given pbuf. Noone</FONT>
<BR><FONT SIZE=2 FACE="Arial">&nbsp;&nbsp; seems to be free-ing those.</FONT>
</P>

<P><FONT SIZE=2 FACE="Arial">So, am I correct that some memory may be leaking 
here, or am I really</FONT>
<BR><FONT SIZE=2 FACE="Arial">missing something?</FONT>
</P>

<P><FONT SIZE=2 FACE="Arial">Paul</FONT>
</P>

</BODY>
</HTML>
------_=_NextPart_001_01C1C11F.BD106600--

--------------InterScan_NT_MIME_Boundary--

[This message was sent through the lwip discussion list.]
reply via email to
[Prev in Thread] Current Thread [Next in Thread]