buglwIP - A Lightweight TCP/IP stack - Bugs: bug #39124, fragmented ICMP output corrupted...

 
 

You are not allowed to post comments on this tracker with your current authentication level.

bug #39124: fragmented ICMP output corrupted by IGMP input

Submitted by:  beaujouan <rico37>
Submitted on:  Thu 30 May 2013 01:52:08 PM UTC  
 
Category: IPv4Severity: 3 - Normal
Item Group: Faulty BehaviourStatus: Fixed
Privacy: PublicAssigned to: Simon Goldschmidt <goldsimon>
Open/Closed: ClosedPlanned Release: None
lwIP version: 1.3.1

(Jump to the original submission Jump to the original submission)

Thu 26 Feb 2015 09:34:15 PM UTC, comment #6:

Without getting more input, this looks too much like a threading issue in the port to leave it open as a bug in core lwIP code.

Simon Goldschmidt <goldsimon>
Project AdministratorIn charge of this item.
Mon 01 Jul 2013 01:05:28 PM UTC, comment #5:

based on what you said, conditions for having a thread safe pbuf alloc/free are met, so yes, it is ! :)

Sylvain Rochet <gradator>
Project Member
Mon 01 Jul 2013 12:59:34 PM UTC, comment #4:

Thanks for your comment Sylvain,

SYS_LIGHTWEIGHT_PROT is set to 1 and sys_arch_protect/sys_arch_unprotect are defined disabling/enabling the MCU global interrupts.

In fact we are using the sys_arch port from Texas Instrument for LM3S9D96 MCU.

So regarding your comments, the pbuf is really thread safe, isn't it ?

beaujouan <rico37>
Mon 01 Jul 2013 12:44:09 PM UTC, comment #3:

pbuf alloc/free operations are thread safe if SYS_LIGHTWEIGHT_PROT is set and sys_arch_protect() / sys_arch_unprotect() defined.

Sylvain Rochet <gradator>
Project Member
Mon 01 Jul 2013 11:55:56 AM UTC, comment #2:

We are using lwIP in a multithreaded environnment (SafeRTOS) so NO_SYS is defined to 0.

I thought pbuf management was thread safe.
How can I be sure that it is due to a multi thread issue ?

beaujouan <rico37>
Fri 28 Jun 2013 08:04:42 PM UTC, comment #1:

Are you absolutely sure this isn't a threading problem? I do not undertand why and incoming packet should corrupt an outgoing packet when lwIP is not multithreaded.

Simon Goldschmidt <goldsimon>
Project AdministratorIn charge of this item.
Thu 30 May 2013 01:52:08 PM UTC, original submission:

Hello,

I am new to savannah, hoping I post in the right place.

First, bug reports to lwIP 1.3.2 but it was not listed in lwIP version list, so I selected 1.3.1.

The bug is, when lwIP send a fragmented ICMP Echo reply, this output frame can be corrupted by an IGMP incoming request.

I have seen the bug #35616 report which seems to be very similar but does not concern lwIP 1.3.2.

According to this bug, in my configuration, ETH_PAD_SIZE = 2 and MEM_ALIGNMENT=4, so I am not sure that issue is the same as in the bug #35616

I think there may be a freeing issue, allowing pbuf used in output frame to be used when receiving the IGMP incoming frame but I am a little bit lost in the pbuf mecanism.

Is there a known fix I did not catch ?
Is there some other information that can I provide to help understanding ?

Regards,

beaujouan <rico37>

 

No files currently attached

 

Depends on the following items: None found

Items that depend on this one: None found

 

Carbon-Copy List
  • -unavailable- added by gradator (Posted a comment)
  • -unavailable- added by goldsimon (Posted a comment)
  • -unavailable- added by rico37 (Submitted the item)
  •  

    Do you think this task is very important?
    If so, you can click here to add your encouragement to it.
    This task has 0 encouragements so far.

    Only logged-in users can vote.

     

    Please enter the title of George Orwell's famous dystopian book (it's a date):

     

     

    Follow 4 latest changes.

    Date Changed By Updated Field Previous Value => Replaced By
    Thu 26 Feb 2015 09:34:15 PM UTCgoldsimonStatusNeed Info=>Fixed
      Assigned toNone=>goldsimon
      Open/ClosedOpen=>Closed
    Fri 28 Jun 2013 08:04:42 PM UTCgoldsimonStatusNone=>Need Info

    Back to the top


    Powered by Savane 3.1-cleanup1