Skip site navigation (1)Skip section navigation (2)
Date:      Thu, 22 May 2003 17:37:02 +0200 (CEST)
From:      Dirk-Willem van Gulik <dirkx@webweaving.org>
To:        FreeBSD-gnats-submit@FreeBSD.org
Subject:   i386/52581: Boot loaders reading more than one sector at the time
Message-ID:  <200305221537.h4MFb2Tt053989@foem.leiden.webweaving.org>
Resent-Message-ID: <200305221540.h4MFe54J053314@freefall.freebsd.org>

next in thread | raw e-mail | index | archive | help

>Number:         52581
>Category:       i386
>Synopsis:       Boot loaders reading more than one sector at the time
>Confidential:   no
>Severity:       serious
>Priority:       high
>Responsible:    freebsd-i386
>State:          open
>Quarter:        
>Keywords:       
>Date-Required:
>Class:          sw-bug
>Submitter-Id:   current-users
>Arrival-Date:   Thu May 22 08:40:05 PDT 2003
>Closed-Date:
>Last-Modified:
>Originator:     Dirk-Willem van Gulik
>Release:        FreeBSD 5.0-RELEASE i386
>Organization:
>Environment:
System: FreeBSD foem.leiden.webweaving.org 5.0-RELEASE FreeBSD 5.0-RELEASE #0: Mon Apr 28 16:12:26 CEST 2003 dirkx@foem.leiden.webweaving.org:/usr/obj/usr/src/sys/FOEM i386

i386 machines booting from recent CompactFlash ROM's from a range
of vendors; including SanDisk which OEM's for a very large number
of others which have BIOS-es which do not nicely go to 'read one
sector at a time' on error.

>Description:
	http://webfolder.wleiden.webweaving.org/apacer-flash/	
		msg2.txt

Essentially an increasing number of vendors sells flash
cards which cannot cope wiht a read of more than one sector
at the time.

>How-To-Repeat:
	Requires hardware booting from flash and a broken
	flash and a BIOS which is too simple. Soekris boxes are
	a good example. And more and more of the flash coming _new_
	on the market (i.e. the larger sizes; and not the so called
	faster flash) has this issue.
>Fix:
	http://webfolder.wleiden.webweaving.org/apacer-flash/ 
	I.e. force the reads to be one sector at the time; makes
	the boot 0.5-1.5 second longer on the hardware tried.

	Or nice; detect the error and then go back to 1-at-the-time
	reading.

	But given the small, if any, speed penalty, this may not
	be actually needed.

>Release-Note:
>Audit-Trail:
>Unformatted:



Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?200305221537.h4MFb2Tt053989>