This is the mail archive of the 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]

[Bug 1001837] Rich FlexBus RAM layout

Please do not reply to this email, use the link below.

--- Comment #16 from Mike Jones <> ---
Created attachment 2227
FXM Specific Exception

This test case will generate an exception on FXM hardware that cannot be
generated on non-FXM hardware. It manifests like the description in Bug 1001764
- Enhancement of MMC/SD over SPI driver. The zip contains an ecm file, code,
and all patches used.

When the application is run, a loop will open a file on an SD card, develop
strings with snprintf, write them to the file, flush the filesystem, and close
the file. When it is run, the problem always occurs when the variable "dev" in
the SPI code is 0xFFFFFFFF and results in an exception 5. It has to be run for
many minutes to cause failure, but typically it will fail in less than 10 mins.


If you have time to run this, it might eliminate or confirm if the problem is
related/unrelated to hardware reliability. Once we know it is not related to
bad hardware, perhaps we can form some theories about the cause of the

Please note that all I2C code is commented out, so this only uses FXM and SD
card and should run on any FXM hardware with SD support. I ran it on a K60
100Mhz, so there is no cache. I have disabled the cache in the ecm as well.

I put the attachment here rather than the SPI bug because I cannot reproduce it
on non-FXM hardware. If it really is a MMC bug, we can move it.

You are receiving this mail because:
You are the assignee for the bug.

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