Bug 27821

Summary:	Redboot mangling dependent on flash driver?!
Product:	[Retired] eCos	Reporter:	Jonathan Larmour <jlarmour>
Component:	RedBoot	Assignee:	ecc-bugs-int
Status:	CLOSED CURRENTRELEASE	QA Contact:	ecc-bugs-int
Severity:	medium	Docs Contact:
Priority:	medium
Version:	CVS
Target Milestone:	---
Target Release:	---
Hardware:	All
OS:	Linux
Whiteboard:
Fixed In Version:		Doc Type:	Bug Fix
Doc Text:		Story Points:	---
Clone Of:		Environment:
Last Closed:	2001-03-07 08:33:05 UTC	Type:	---
Regression:	---	Mount Type:	---
Documentation:	---	CRM:
Verified Versions:		Category:	---
oVirt Team:	---	RHEL 7.3 requirements from Atomic Host:
Cloudforms Team:	---	Target Upstream Version:
Embargoed:

Description Jonathan Larmour 2001-02-15 18:11:17 UTC

Look at the thread "Mangling output" on Thursday 15th Feb 2001. RedBoot
wasn't mangling console output, then Hugo found this out:

"
I persisted with my other work - adding a flash driver - 'cos it needs
doing anyway, and doing that brings my target more into line with the other
targets that work.  And LO! suddenly mangling works too.

Yes, just confirmed that.  Build RedBoot but remove the flash packages and
it fails, build it "vanilla" and it works, the very same app binary and
everything - just a new ROM build with FLASH packages.

So, there's a RedBoot bug in there somewhere, I suspect, but I don't care -
it works now.

"

Comment 1 Jesper Skov 2001-02-22 08:41:02 UTC

Jesper> OK, I'll have a look at the CDL state.

Uh, baffled. No changes in the pkgconf headers (or CDL) which should
be even remotely related.

Off the top of my head, I'd guess at something related to
caching. Flash drivers do some flushing at init time which you don't
get otherwise.  And looking at the sa11x0 init code, there's no flush
before caches are enabled... Guess it could be done earlier, but
hal_if_init call is just before caches get enabled...

Jesper

Comment 2 Jesper Skov 2001-03-07 08:32:55 UTC

Was due to missing HAL_DIAG_INIT call - now done from RedBoot main function.
(including flash drivers would pull in diag_printf clients and thus the init
call).