[Helix-client-dev] Helix build on Linux - query

[Helix-client-dev] Helix build on Linux - query

Jagmeet Singh jsingh at real.com
Tue Aug 8 07:56:46 PDT 2006


Hi Greg,

Has there been any bug filed for this issue?

Regards,
Jagmeet Singh
Software Development Engineer
Helix Community
Real Networks, Inc. 
jsingh at real.com 
http://www.helixcommunity.org
http://www.realnetworks.com/products/support/devsupport.html


-----Original Message-----
From: Greg Wright [mailto:gwright at real.com] 
Sent: Tuesday, August 01, 2006 11:12 PM
To: jsingh at real.com
Cc: helix-client-dev at lists.helixcommunity.org; Anil Jaiswal
Subject: Re: [Helix-client-dev] Helix build on Linux - query

Jagmeet Singh wrote:
> Hi Greg,
> 
> Did you get a chance to look into this behavior? I have attached the
> latest call stack for the crash incase it helps.
> 
> I would appreciate if I could get some directions as to how I may
> proceed. Given below are the buffer overrun stacks reported by
valgrind
> for reference.

Anil is currently working on this. He will be filing a bug
so that you can track its progress. Anil, please respond to
this email with the bug you create.

--greg.


> 
> ==15497==
> ==15497== Invalid write of size 1
> ==15497==    at 0x401C5B7: memcpy (mac_replace_strmem.c:394)
> ==15497==    by 0x48AB7BB: CHXBuffer::Set(unsigned char const*,
unsigned
> long) (hxbuffer.cpp:315)
> ==15497==    by 0x48AB195: CHXBuffer::CHXBuffer(unsigned char*,
unsigned
> long, int) (hxbuffer.cpp:137)
> ==15497==    by 0x48A24A6:
> CHXBufferFragment::CHXBufferFragment(IHXBuffer*, unsigned char*,
> unsigned long) (hxbuffer.h:223)
> ==15497==    by 0x489CFA5:
NewPacketBufferGetter(IHXCommonClassFactory*,
> IHXBuffer*, unsigned char*, int) (mp3ff.cpp:107)
> ==15497==    by 0x489EFCB: CRnMp3Fmt::MyCreatePacketObj_hr(long,
> CRnMp3Fmt::tReadBuffer*) (mp3ff.cpp:1331)
> ==15497==    by 0x48A1B35: CRnMp3Fmt::FinishGetPacket()
(mp3ff.cpp:2748)
> ==15497==    by 0x489EC78: CRnMp3Fmt::GetPacket(unsigned short)
> (mp3ff.cpp:1103)
> ==15497==    by 0x4E8E34E:
> CHXPlaybackVelocityProxyFF::_GetPacket(CHXStreamInfo*)
> (velproxy.cpp:2046)
> ==15497==    by 0x4E8BD90:
> CHXPlaybackVelocityProxy::ExternalGetPacket(unsigned long)
> (velproxy.cpp:895)
> ==15497==    by 0x4E8DF50:
> CHXPlaybackVelocityProxyFF::GetPacket(unsigned short)
> (velproxy.cpp:1928)
> ==15497==    by 0x4DA1CDC: HXFileSource::FillBuffers()
> (hxflsrc.cpp:3239)
> ==15497==    by 0x4D9EAFD: HXFileSource::GetEvent(unsigned short,
> CHXEvent*&) (hxflsrc.cpp:1745)
> ==15497==    by 0x4D8DC6A: SourceInfo::ProcessIdle(int, unsigned
long&,
> int&, unsigned short&) (srcinfo.cpp:1281)
> ==15497==    by 0x4D688CE: HXPlayer::ProcessIdle() (hxplay.cpp:2264)
> ==15497==    by 0x4D7461F: HXPlayer::PlayerCallback(void*)
> (hxplay.cpp:7786)
> ==15497==    by 0x4E5A5EF: CHXGenericCallback::Func()
(hxcbobj.cpp:103)
> ==15497==    by 0x44D09B2: PQ::dispatch_element(PQElem*) (pq.cpp:465)
> ==15497==    by 0x44C6E8F: ClientPQ::execute(Timeval)
(clientpq.cpp:112)
> ==15497==    by 0x4494450: HXScheduler::ExecuteCurrentFunctions(int)
> (hxsched.cpp:441)
> ==15497==    by 0x4494EC8: HXScheduler::OnTimeSync(int)
> (hxsched.cpp:768)
> ==15497==    by 0x44652BD: CHXMediaPlatformScheduler::OnTimeSync(int)
> (chxmedpltfmsched.cpp:160)
> ==15497==    by 0x44646CA: CHXMediaPlatformKicker::Kick(unsigned long)
> (chxmedpltfmkicker.cpp:275)
> ==15497==    by 0x804AC08: DoEvent() (main.cpp:423)
> ==15497==    by 0x804E9B3: main (main.cpp:2065)
> ==15497==  Address 0x4533320 is not stack'd, malloc'd or (recently)
> free'd
> 
> valgrind: m_mallocfree.c:186 (get_bszB_as_is): Assertion 'bszB_lo ==
> bszB_hi' failed.
> valgrind: Heap block lo/hi size mismatch: lo = 81, hi = 1527556001.
> Probably caused by overrunning/underrunning a heap block's bounds.
> 
> Regards,
> Jagmeet Singh
> Software Development Engineer
> Helix Community
> Real Networks, Inc. 
> jsingh at real.com 
> http://www.helixcommunity.org
> http://www.realnetworks.com/products/support/devsupport.html
> 
> 
> -----Original Message-----
> From: Jagmeet Singh [mailto:jsingh at real.com] 
> Sent: Wednesday, July 26, 2006 11:15 AM
> To: 'Greg Wright'
> Cc: 'helix-client-dev at lists.helixcommunity.org'
> Subject: RE: [Helix-client-dev] Helix build on Linux - query
> 
> Sorry, missed the callstack.
> 
> Attached.
> 
> Mostly I get the segmentation fault at this place. 
> 
> 
> Jagmeet Singh
> Software Development Engineer
> Helix Community
> Real Networks, Inc. 
> jsingh at real.com 
> http://www.helixcommunity.org
> http://www.realnetworks.com/products/support/devsupport.html
> 
> 
> -----Original Message-----
> From: Greg Wright [mailto:gwright at real.com] 
> Sent: Wednesday, July 26, 2006 5:18 AM
> To: jsingh at real.com
> Cc: helix-client-dev at lists.helixcommunity.org
> Subject: Re: [Helix-client-dev] Helix build on Linux - query
> 
> I don't see the call stack.
> 
> --greg.
> 
> 
> 
> Jagmeet Singh wrote:
>> Hi Greg,
>>
>> I am getting segmentation faults with the mp3 content only, however
> the
>> rm, 3gp and mp4 contents play without any issues. I tried using the
>> nullrenderer to rule out the mp3 renderer issues.
>>
>> I used gdb to debug the issue and got the call stack which is
attached
>> to this mail. The crash is happening mostly in the new operator call.
>> Lately now sometimes I do get a crash in the free operator calls as
>> well.
>>
>> I have tried using valgrind to check for memory corruption but so far
>> didn't get much from that either.
>>
>> I did a diff of the mp3 fileformat files (from hxclient_1_5_0_cayenne
>> branch) as well to check for any major changes that have gone into it
>> but did not come across any that would result in something like this.
>>
>> I have tried mp3fformat.so plugin with the hxclient_1_5_0_cayenne
> branch
>> as well and the mp3 files played without any problems.
>>
>> Regards,
>> Jagmeet Singh
>> Software Development Engineer
>> Helix Community
>> Real Networks, Inc. 
>> jsingh at real.com 
>> http://www.helixcommunity.org
>> http://www.realnetworks.com/products/support/devsupport.html
>>
>>
>> -----Original Message-----
>> From: Greg Wright [mailto:gwright at real.com] 
>> Sent: Wednesday, July 19, 2006 10:43 PM
>> To: jsingh at real.com
>> Cc: helix-client-dev at lists.helixcommunity.org
>> Subject: Re: [Helix-client-dev] Helix build on Linux - query
>>
>> Jagmeet Singh wrote:
>>> Hi,
>>>  
>>> I am working on the helix-restricted build on Linux. I am
> experiencing
>>> some problems with the local playback of mp3 content, as well as
with
>>> the streaming of all the media formats. I wanted to confirm if this
>>> branch has been tested on the Linux platform. The source build on
>>> windows is working fine in the same scenarios.
>> We build on the helix-restricted branch nightly but I don't believe
>> any testing is being done on it. So, it is quite possible that there
>> might be a runtime problem we do not know about. Can you be more
>> specific on what problems you are having any any debugging you have
>> done so far?
>>
>> There have been lots of changes on HEAD for the new Atlas platform
>> and I don't know how well they have been tested on non-window's
>> platforms at this point.
>>
>> --greg.
>>
>>




More information about the Helix-client-dev mailing list
 

Site Map   |   Terms of Use   |   Privacy Policy   |   Contact Us

Copyright © 1995-2007 RealNetworks, Inc. All rights reserved. RealNetworks and Helix are trademarks of RealNetworks.
All other trademarks or registered trademarks are the property of their respective holders.