This is the mail archive of the ecos-discuss@sourceware.org mailing list for the eCos project.


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]
Other format: [Raw text]

Re: RE: Problem with FAT32


Himanshu Patel wrote:
> Hi Gary,
> 
> 'fatfs1.c' is the same as the standard one. It does fail when we have other
> thread with IO operation (for example accessing UART).

Then it's not the same as the standard one, is it?  The file you sent
is not a stand alone test showing your problem.  I also note that your
modified test accesses a MMC/SD drive - does the problem still happen
if you use some other disk, for example a RAM disk?

Either show us how to duplicate your failure, e.g. how you modified
that test to add additional threads, how to run it, etc, or stop bothering
this public forum.

> -----Original Message-----
> From: Gary Thomas [mailto:gary@mlbassoc.com] 
> Sent: Friday, March 20, 2009 7:34 PM
> To: Himanshu Patel
> Cc: 'eCos Discuss'
> Subject: Re: [ECOS] RE: Problem with FAT32
> 
> Himanshu Patel wrote:
>> Hi Gary,
>>
>> Please find my reply:
>>   * What [else] have you done to isolate the problem?
>>     [Himanshu] 
>>     1. We tried to integrate the file system operations with other thread
>> (now we have only one thread) and system is 
>>        running okay.
>>     2. We tried to increase the thread stack size and interrupt stack
> size.
>> But the results are the same. 
>>
>>   * Are you running with asserts (and checks) enabled?
>>     [Himanshu] Yes...We have asserts enabled. But reset is not happening
> due
>> to assert. 
>>
>>   * Can you duplicate it outside of your application?
>>     [Himanshu] We tried to run two simple threads (one doing file
> operation
>> and other doing writing on UART). Then also it     
>>     resets.  
> 
> Isolate this test.  Does it only fail when using MMC/SD access?
> 
>>   * Can you share [any of] the code that fails?
>>     [Himanshu] Okay...Please find herewith application code for your
>> reference. Note that this is very simple 
>>     application which we are using right now for basic testing.
> 
> Not terribly useful, frankly.  I'm not interested in reading 1000 lines
> of random code without any guidance, etc.  For example, how does your
> 'fatfs1.c' differ from the standard one?  Does it fail as well?
> 
>> Let me know if you need any further information which would help to
> identify
>> the problem/solution.
> 
> What you'll get here is at best _guidance_, not analysis and debug.
> As Chris said, that's what *paid* support is for.
> 
>> -----Original Message-----
>> From: Gary Thomas [mailto:gary@mlbassoc.com] 
>> Sent: Friday, March 20, 2009 4:39 PM
>> To: Himanshu Patel
>> Cc: 'eCos Discuss'
>> Subject: Re: [ECOS] RE: Problem with FAT32
>>
>> Himanshu Patel wrote:
>>> Gentle reminder...We are kind of stuck and delivery date is very near.
>> Gentle reply - this is a volunteer list, we'll help _when_ we can.
>>
>> That said, more details are probably necessary in order to be able
>> to make any comments.
>>   * What [else] have you done to isolate the problem?
>>   * Are you running with asserts (and checks) enabled?
>>   * Can you duplicate it outside of your application?
>>   * Can you share [any of] the code that fails?
>>
>>> -----Original Message-----
>>> From: ecos-discuss-owner@ecos.sourceware.org
>>> [mailto:ecos-discuss-owner@ecos.sourceware.org] On Behalf Of Himanshu
>> Patel
>>> Sent: Thursday, March 19, 2009 8:12 PM
>>> To: eCos Discuss
>>> Subject: [ECOS] RE: Problem with FAT32
>>>
>>> Note that failure generally happens when we have opendir() or readdir()
>>> calls. Any one has encountered such problem?
>>>
>>> Regards,
>>>
>>> Himanshu Patel
>>>
>>> -----Original Message-----
>>> From: Himanshu Patel 
>>> Sent: Thursday, March 19, 2009 6:57 PM
>>> To: 'eCos Discuss'
>>> Subject: Problem with FAT32
>>>
>>> Hi,
>>>
>>> We are using FAT32 file system in our project. 
>>>
>>> We are facing strange problem. When we test the file system commands
>>> (open, close, read, write) all works very correctly in single thread.
>>> However when we invoke other thread, then file system commands do not
>>> work and board gets reset. 
>>>
>>> We have allocated thread stack memory of  file system thread up to 10 KB
>>> and for other thread is 4 KB. 
>>>
>>> We have tried to combine both the threads  in one thread and then also
>>> the results are the same (board gets reset). 
>>>
>>> We have tried to find out problem by enabling asserts and tracing.
>>> However we don't get any assert at the time of reset. 
>>>
>>> What could be the issue? 
>>>
>>> Regards,
>>>
>>> Himanshu Patel
>>>
>>
> 
> 


-- 
------------------------------------------------------------
Gary Thomas                 |  Consulting for the
MLB Associates              |    Embedded world
------------------------------------------------------------

-- 
Before posting, please read the FAQ: http://ecos.sourceware.org/fom/ecos
and search the list archive: http://ecos.sourceware.org/ml/ecos-discuss


Index Nav: [Date Index] [Subject Index] [Author Index] [Thread Index]
Message Nav: [Date Prev] [Date Next] [Thread Prev] [Thread Next]