This is the mail archive of the
ecos-patches@sources.redhat.com
mailing list for the eCos project.
RE: EB40a evaluation LEDs
- From: Øyvind Harboe <oyvind dot harboe at zylin dot com>
- To: "'Jonathan Larmour'" <jifl at eCosCentric dot com>, "'Andrew Lunn'" <andrew dot lunn at ascom dot ch>
- Cc: <ecos-patches at sources dot redhat dot com>
- Date: Mon, 16 Jun 2003 08:58:09 +0200
- Subject: RE: EB40a evaluation LEDs
>I don't have hardware here either. Nick does have an EB40a, but to be
>honest, I don't think in general we should rely on one person to test a
>patch as a prerequisite for committing otherwise the patch backlog will
>grow even quicker! Of course if Nick or someone else wants to test and
>report back, then the patch can easily be reverted if need be.
>
>But I'm sure most people who write these patches do so for a reason and
do
>test them! Probably not always, but hopefully enough.
The GCC compiler guys have a list of configurations where the GCC
compiler
has been reported to work. http://gcc.gnu.org/buildstat.html
Does a similar list exist for eCos?
Something like the results of:
1. Compile Redboot RAM/ROM/ROMRAM
2. Load an app in Redboot RAM/ROM/ROMRAM. ("smoketest" would be a
suitable
name).
3. Use Redboot RAM/ROM/ROMRAM to write an app to flash (+ booting the
app).
Clearly it is impossible to test all permuations of eCos. Even with a
single
hardware target, you still have the .cdl configuration switches that
create
FAPP an infinite number of combinations.
Which is why open source development and a system for submitting patches
is *great*, BTW. :-)
Øyvind