Opened 10 years ago

Closed 6 years ago

#261 closed defect (Unsupported Version)

USB drive problem

Reported by: losepete Owned by: eco
Priority: major Milestone: Release version 3.19
Component: ACPI PSD Version: 3.10
Keywords: Cc:

Description

After applying ACPI support to a working eCS1.2R system the system "lost" the ability to have more than 1 USB drive attached at the same time; if data/applications are required from a different USB drive it is necessary to Eject the currently attached USB drive before attaching a different USB drive.

Having performed a clean, minimal eCS2.0RC5 installation *Without* ACPI support during the installation and *Without* selecting to install ACPI I can state there was not a USB drive problem ie: No ACPI involved, No USB drive problem - I can attach all 4 USB drives available here *at the same time*

I then reinstalled eCs2.0RC5 with ACPI support, installed and configured ACPI support.

I corrected the ACPI Wizard's failure to add the "/!NOD" switch that I specified for the config.sys line and rebooted. PSD line in use is :-

PSD=ACPI.PSD /SMP /APIC /TMR /CD /!NOD

After the system had booted it was no longer possible to attach more than 1 USB drive and get the system to recognise the 2nd USB drive. The 2nd USB drive LED flashes when plugged in but the usual "Device x: has been attached" message does not appear, the drive is not in the Drives folder, Refresh Removable Drives does not cause it to appear - it is seen by tools such as USBDock though.

Regards

Pete

Attachments (2)

acpi-log-Model-of-your-PC.ziq (70.3 KB) - added by losepete 10 years ago.
*.ziq file created by the acpi logs collector
config.sys (8.2 KB) - added by losepete 10 years ago.

Download all attachments as: .zip

Change History (35)

comment:1 Changed 10 years ago by eco

e-co:

1) model of your computer? BIOS setup settings related to USB?

2) please, show us some logs: http://ecomstation.ru/projects/acpitools/?action=logs

3) attach CONFIG.SYS

Changed 10 years ago by losepete

*.ziq file created by the acpi logs collector

Changed 10 years ago by losepete

comment:2 Changed 10 years ago by losepete

Asus M2N am2 mainboard, nForce430/MCP61 chipset

BIOS settings

Plug'n'Play OS: Yes USB1.1 Controller: Enabled USB2 Controller: Enabled Legacy USB support: Disabled USB2 Controller mode: HiSpeed? (480Mbps)

I updated ACPI to v3.10 without any change in the USB drive problem - the collected logs in the *.ziq file are all after the update.

Should I open another ticket to report the lack of (eStyler) Shutdown reaction in ACPI v3.0x? - sometimes Power Off actually works but usually clicking the eComCenter Shutdown and selecting from Power Off/Shutdown/Reboot? results in nothing happening.

Usually the 1st click of the Shutdown button Power off has the Shutdown dialog offering Power Off(APM). If I click Cancel and then click Shutdown again that dialog will change to Power Off (ACPI).

On my eCS1.2R system I usually use shutdown.exe to get the system to shutdown.

11-12-06 5:50p 11,776 342 a--- shutdown.exe

Regards

Pete

comment:3 Changed 10 years ago by eco

e-co:

1) stop.. USB dock can see devices attached to USB?

this is a problem of usb flash disks, start from this message: http://ecomstation.ru/commentnews.php?id=1520

2) shutdown - don't use eStyler, use eWorkPlace shutdown http://ecomstation.ru/projects/acpitools/Shutdown

comment:4 Changed 10 years ago by losepete

1) Yes, USBDock can see a 2nd (and 3rd) USB drive attached to the system but not useable by the system.

Your url leads to the "ACPI 3.10 is available" announcement... As this report covers v3.10 as well I'm not sure what you are trying to indicate

2) eWorkPlace *Not* Installed so that is not an option. The shutdown.exe mentioned previously does work and seems to be my only option apart from WatchCat? or C-A-D.

comment:5 Changed 10 years ago by pasha

  • Version changed from 3.09 to 3.10

Remove /!NOD from apci.psd and do new acpi.log pls

comment:6 Changed 10 years ago by eco

Hello losepete

1) OK, so, ACPI works fine: the devices are visible => USB controllers are initialized and available for the OS.

i don't know, does it help or not, try increase the value of removables: CONFIG.SYS -> BASEDEV=USBMSD.ADD /REMOVABLES:12 /FLOPPIES:0

comment:7 Changed 10 years ago by losepete

Hi Pasha

My attempts at booting without the /!NOD switch have not been successful - this is an nVidia nForce430/MCP61 chipset.

By a process of REMming a few drivers I find that the driver that takes a long while - several minutes - to load is DEVICE=L:\ECS\SYSTEM\AMouse\USBMOUSE.SYS

After that USB Mouse driver has loaded the boot process slows down and I think it actually stops - after 30 minutes I decided it must have given up and rebooted to eCS1.2R to post this result.

So, no log; not possible to get to the position of being able to create the log.

comment:8 follow-up: Changed 10 years ago by pasha

Try use IBM usbmouse?

comment:9 in reply to: ↑ 8 Changed 10 years ago by losepete

Replying to pasha:

Try use IBM usbmouse?

Hi Pasha

I uninstalled AMouse and the uninstall routine reinstated the IBM Mouse files.

I rebooted and all was going well until the (IBM) USBMouse driver was loading. The system took a while to display that driver had been loaded and then took about 5 minutes for each config.sys entry before finally getting stuck - I presume it was stuck as after 12 minutes there had been no further driver load messages and there were still a few to come.

As I mentioned previously the /!NOD switch is *Necessary* with every ACPI v3.xx package that I have tried and v3.10 is no different.

comment:10 Changed 10 years ago by pasha

Why you are use /!NOD ? Can you boot without /!NOD and attach log?

comment:11 Changed 10 years ago by losepete

Hi Pasha

I use /!NOD to get the system to boot

No, I cannot boot the system without /!NOD

comment:12 Changed 10 years ago by pasha

Where boot is stopping? without /!NOD

comment:13 Changed 10 years ago by losepete

Hi Pasha

This is getting a bit circular...

When booting without /!NOD booting slows down a lot when the USB Mouse driver is loading and stops shortly after.

It takes about 5 minutes to load the USB Mouse driver and then every driver after that takes longer and longer until the system simply stops trying to boot - maybe it would boot if I left it alone for a day or several...

This is true with both AMouse and IBM drivers.

comment:14 Changed 10 years ago by pasha

Give me log after this boot pls

comment:15 Changed 10 years ago by losepete

Hi Pasha

This is probably going to sound rather strange but:

Having uninstalled AMouse and installed the IBM Mouse drivers, added /NOD to the PSD= line I could not boot to a working Desktop, the system displayed the "An Internal Processing Error has occurred message". I REM'd the IBM Mouse driver lines, booted to a "mouse free" Desktop and re-installed the AMouse driver supplied with eCS2.0RC5

I then booted and discovered that the "only 1 USB drive at a time" problem is gone!

Yes, I can now attach 3 USB drives at the same time - no more currently here to test - and all it seems to be the result of uninstalling AMouse, installing the IBM Mouse (done when using Warpin to uninstall AMouse) and then reinstalling AMouse.

Go Figure!

The AMouse.sys driver does not have the SMP switch and seems to be working fine without it.

Could it be something to do with driver order in the config.sys file? - although I did not think that mattered as much as back in the early days of Warp3/4. The AMouse driver lines are now near the end of the config.sys file whereas they were previously much nearer the start, in the section eCS installs label [Device Drivers].

A little more testing and experimenting has led to the discovery that the nForce430/MCP61 chipset does *Not* require the use of the /TMR switch with ACPI v3.09/v3.10 - using it screws up the tmrtest.exe results which are fine without it.

I will have another try at booting without /!NOD and post logs if successful.

Regards

Pete

comment:16 Changed 10 years ago by losepete

No, definitely cannot boot without the /!NOD switch

In an earlier response it was mentioned to "increase the value of removables: CONFIG.SYS -> BASEDEV=USBMSD.ADD /REMOVABLES:12 /FLOPPIES:0"

I currently have BASEDEV=USBMSD.ADD /REMOVABLES:8 /FLOPPIES:0 for both eCS2.0RC5/ACPIv3.10 and eCS1.2R/ACPIv3.09 installations.

If I reboot with that line changed to BASEDEV=USBMSD.ADD /REMOVABLES:5 /FLOPPIES:0 the "only 1 USB drive can be attached at a time" problem resurfaces. This is common to both eCS2.0RC5/ACPIv3.10 and eCS1.2R/ACPIv3.09 installations.

Without ACPI installed the problem does not exist so I am guessing that ACPI gets something wrong with regard to USB drives...

comment:17 Changed 10 years ago by losepete

Hi Pasha

Is it possible that using ACPI causes the system to recognise my 4 slot Internal (attached from boot) USB cardreader and therefore has 4 of the /REMOVABLES: value already "reserved" for those 4 slots?

It would be a very neat explanation as to why /REMOVABLES:5 will only allow 1 USB hard/flash drive to be attached at a time and /REMOVABLES:8 will allow all 3 available to be attached.

As mentioned previously: Without ACPI installed there is no problem with the /REMOVABLES:5 value; it is only after installing ACPI that the problem occurs.

comment:18 Changed 10 years ago by losepete

Hi Pasha

Updated ACPI to V3.11

Removed "/!NOD" from the config.sys PSD= line and rebooted.

Same problem as with V3.10; system stops booting after loading USB drivers.

Looks like the "/!NOD" switch is still required for nVidia nForce430/MCP61 based systems.

Regards

Pete

comment:19 Changed 9 years ago by pasha

Need try and give log with 3.14

comment:20 Changed 9 years ago by pasha

Please download experimental ACPI build from Mensys site:

  • Experimental build for you:

ACPI-APIC-BAT.ZIP

comment:21 Changed 9 years ago by stevenhl

  • Milestone changed from Release version 3.15 to Feedback pending

comment:22 Changed 9 years ago by pasha

  • Owner pasha deleted

comment:23 Changed 9 years ago by losepete

Currently running eCS2.0RC6a which comes with v3.14

This works fine with the following config.sys line:-

PSD=ACPI.PSD /SMP /CD /TMR /APIC /!NOD

If I change that line to

PSD=ACPI.PSD /SMP /CD /TMR /APIC

and reboot the boot gets stuck at USBEHCD.SYS

I have installed and tested both ACPI-APIC-BAT.ZIP files and ACPI315PRE.ZIP files as well.

Both fail to boot if the config.sys line is more than this:-

PSD=ACPI.PSD

Not much use on a dual or multi core system...

If I try this line:-

PSD=ACPI.PSD /SMP /CD /TMR /APIC /!NOD

The boot stops - see Ticket #392 ACPI315PRE fails to boot for details as it applies to both ACPI-APIC-BAT.ZIP and ACPI315PRE.ZIP packages.

As with most earlier versions of ACPI if I remove the /!NOD switch bootup stops at USBEHCD.SYS

Conclusion: v3.14 works better than the later versions.

comment:24 follow-up: Changed 9 years ago by mckinnis

Did you check my notes on this ticket?<br> http://svn.netlabs.org/acpi/ticket/391#comment:5

comment:25 in reply to: ↑ 24 ; follow-up: Changed 9 years ago by losepete

Replying to mckinnis:

Did you check my notes on this ticket?<br> http://svn.netlabs.org/acpi/ticket/391#comment:5

I think you may have posted to the wrong ticket - or I am missing something?

comment:26 in reply to: ↑ 25 ; follow-up: Changed 9 years ago by mckinnis

Replying to losepete:

Replying to mckinnis:

Did you check my notes on this ticket?<br> http://svn.netlabs.org/acpi/ticket/391#comment:5

I think you may have posted to the wrong ticket - or I am missing something?

Both of these tickets seem to be USB related.

comment:27 in reply to: ↑ 26 Changed 9 years ago by losepete

Replying to mckinnis:

Replying to losepete:

Replying to mckinnis:

Did you check my notes on this ticket?<br> http://svn.netlabs.org/acpi/ticket/391#comment:5

I think you may have posted to the wrong ticket - or I am missing something?

Both of these tickets seem to be USB related.

Hi Chuck. More of an ACPI/USB relationship. The USB bit was sorted about 6 months ago; using ACPI seems to mean that my *Internal* USB card reader gets all 4 "drive" slots reserved meaning I could only attach 1 external USB drive when using the USBMSD switch /REMOVABLES:5 That was resolved by changing to /REMOVABLES:8

comment:28 Changed 8 years ago by losepete

I see this still shows as "feedback pending" - even though the feedback was supplied years ago.

comment:29 Changed 8 years ago by pasha

  • Owner set to pasha

One of yours tickets has next:

Please download experimental ACPI build from Mensys site:

  • Experimental build for you:

acpi319ForDesktop.zip

comment:30 Changed 8 years ago by stevenhl

  • Milestone changed from Feedback pending to Release version 3.19

Note to Pete. When a reporter leaves feedback, the reporter is supposed to change the milestone back to the current development milestone. See

TicketProcessingGuide#UsingtheFeedbackPendingMilestone

for more details. The admins and developers are supposed to help the reporters and change the milestone when the reporters forget, but the world is an imperfect place.

comment:31 Changed 7 years ago by pasha

Please download experimental ACPI build from Mensys site:

  • Experimental build for you:

acpi319PollingFullOnly.zip

comment:32 Changed 7 years ago by pasha

  • Owner changed from pasha to eco

feedback pending -> mensys stage

comment:33 Changed 6 years ago by dazarewicz

  • Resolution set to Unsupported Version
  • Status changed from new to closed

This ticket is closed because 1. unsupported version, 2. unrelated to ACPI.

Note: See TracTickets for help on using tickets.