Tiny Core 3.0 Alpha 3 Testing

[jano]

First, you don't disturb me. I went back to the a-2 thread and you never elaborated on which extension(s) you loaded, what kind of hardware, etc., to explain what might be needed to reproduce the same problem. Was this related to 64-bit kernel? Did you get the same or similar errors booting "base norestore" etc.? What chipset was supported in a-1 but not previously?

You are absolutely right. But I didn't elaborate about all these things because I didn't know I had to, my knowledge of linux is very, very limited and nobody asked me to clarify.... so I'll try to make things better now. Excuse me if I post irrelevant information.
* My computer is a Acer Aspire One, AO751H with an Intel Atom processor Z520 with an ide 160 Gb harddisk. TC 2.10 was able to boot from a CD or a pendrive, but it didn't recognize (find?) the harddisk. In a thread I started on that issue curaga pointed that "it's a poulsbo chip. Unsupported currently in TC 2.x. Unless someone makes an extension, wait for 3.x." So I have been using TC 2.10 and 2.11 booting from a ext2 formatted SD card.
* The tests I have done with  alpha-1, alpha-2 and alpha-3 all have been done booting from the same SD card (formatted as ext2, with persistent home and opt). All three alpha's do find the harddisk now.
* With alpha-1 I haven't found any problem so far and thus I use it now as if it were stable. With alpha-2 I found different boot errors (different in different boots as I mentioned in the alpha-2 thread) while loading the TC Extensions. With alpha-3, also boot errors at the same stage.
* As I think that the system is 32 bits, I have only used 32-bit kernel.
* The extensions I load on boot are: alsa, pcmanfm and vlc. Only these three.
* The bootcodes I use are: quiet vga=788 noswap home=sdb1 opt=sdb1 laptop restore=sdb1/tce max_loop=255 waitusb=5
* I changed waitusb to 10 but it didn't make any difference
* Both in alpha-2 or in alpha-3, after the boot errors, the system finally boots and all three extensions do work, but other things, like mounting a pendrive through cpanel's mount do not work (I guess is that it is related to the errors on booting)
I attach the message after booting with syslog enabled.

[gerald_clark]

I suggest you start clean with a new V3 CD and no USB pen drive.

[jano]

I suggest you start clean with a new V3 CD and no USB pen drive.

Will try that for testing and post the results, but as the netbook has no CD drive, I'll have to revert to pendrive for everyday use.

[roberts]

jano, what happens when you boot with the following boot options?

base norestore text

Perhaps you have the same issue as reported above with Acer Aspire and intel graphics chipset.
Worked in alpha1 but not alpha2 or alpha3 until they used nomodeset boot option. Might give that a try as well.

[jano]

Here are the results:
* Booting from a new alpha-3 CD worked OK (no boot errors)

* Booting from the SD card with base norestore text worked OK (no boot errors either). startx then gave me back the WM

I've done some additional testing, always booting from the SD card :
* With base norestore worked OK (no boot errors either) and direct to the WM
* With base and home=sdb1 opt=sdb1 restore=sdb1/tce worked OK (no boot errors either), direct to the WM, have to manually install the extensions but all install and work OK

 

* Finally, removing the base from the boot options [that is, my first boot configuration], brings back the boot errors. So it seems to me that the problem arises when installing the extensions at boot time.
* Adding nomodeset to the boot options doesn't prevent the boot errors

[lucky13]

Here are the results:
* Booting from a new alpha-3 CD worked OK (no boot errors)

* Booting from the SD card with base norestore text worked OK (no boot errors either). startx then gave me back the WM

I've done some additional testing, always booting from the SD card :
* With base norestore worked OK (no boot errors either) and direct to the WM
* With base and home=sdb1 opt=sdb1 restore=sdb1/tce worked OK (no boot errors either), direct to the WM, have to manually install the extensions but all install and work OK

 

* Finally, removing the base from the boot options [that is, my first boot configuration], brings back the boot errors. So it seems to me that the problem arises when installing the extensions at boot time.
* Adding nomodeset to the boot options doesn't prevent the boot errors

Is your Aspire One like mine with two card readers (left and right side)? In my past experience using Linux on my AA1, the right one has seldom worked (in fact, it caused more freezes during boot than it ever had successful boots) and the left one has been a little better. I've never trusted it enough to ever consider booting from it or even to let it mount it at boot (I have automount disabled in both Windows and Scientific Linux on the AA1) and it's pissed me off that it's so dodgy that I have to boot with a card inserted to be able to use it at all. My work around for both card readers in Linux has been to use a small USB card reader instead or directly connect my camera or whatever device; that way I can read SD/Micro/etc. cards whether I have one inserted at boot or not.

If it has two card readers, have you tried the SD card in both slots? Just the right one or just the left one? If you have a USB card reader/adapter, have you tried using it with the usual bootcodes?

Just to make sure we're on the same page, did you get the application extensions from the 3.0 repository or are those ones you have from the 2.0 repositories?

AFAIK, you should've been able to boot earlier TinyCore versions with that Atom CPU. Have you always tried to use the SD card slot for that or have you tried external readers, USB sticks, etc., and left the card slots on the AA1 out of the mix altogether?

[jano]

My AA1 has only one card reader (in the right side) that has always worked without problems in Windows and Linux. It is the way I boot normally TC. I have tried with a usb card adapters with the same results. Booting from a usb pendrive has worked in the same way as with the SD card (and I prefer the SD card as it doesn't protrude out of the netbook). And both with fat and ext2 filesystems.

With the SD card (or pendrive, CD, etc) I hace been able to boot TC 2.10 and TC 2.11, but they didn't "find" the harddrive, so I couldn't install it to the harddisk.

I'm quite sure the extensions come from the 3.0 repository, as I did a fresh "USB install" (not an upgrade) (which formats the card) to the card from an alpha-1 CD and then downloaded the extensions.

My tests for tonight (as I have to go to work now) are (unless someone comes with different sugestions):
* Make and test a new fresh clean install from alpha-3 CD, reformatting the card and downloading the extensions.
* If that fails again, test booting with only one extension at a time, to see if the boot errors appear while installing a specific extension
* Then try with two extensions ...

Thank you for your comments

[curaga]

Do check both the fs and the card for bad blocks. (install e2fsprogs_apps.tcz, run "e2fsck -c -f /dev/sdx1", replace sdx1 with your device)

Also, it's highly peculiar your errors started from alpha2 when no changes were made to anything related. Signs would point to corruption somewhere.

Please also check the md5sums of all your extensions.

[jur]

Something nice:

My wlan sensitivity is up by 20dB! I can now sit comfortably in the lounge by the fire. The EE in me can't explain how a radio can work better to the tune of a whopping 20dB with a new driver but I'll take it thanks!

Something broken:

My touchpad which had 100% workability before in 2.11 does not  respond properly to  tapping; it sorta works but only "halfway" eg tapping on wbar does not cause the icon to depress but does start the app. I also cannot click-drag with double-tapping action like before. Finally, 2-finger dragging across the pad which is supposed to scroll, no longer works.

dmesg in 2.11 reported a Logitech wheel-mouse, whereas now it reports a Elantech touchpad. Well it is a touchpad but it doesn't work vey well. Is there a driver I should load? Or can I somehow invoke the Logitech setup as before?

[Juanito]

Are you using Xorg-7.4? If so, you could try the synaptics touchpad driver.

[jur]

Are you using Xorg-7.4? If so, you could try the synaptics touchpad driver.

Well unfortunately that didn't work so well. Mounted the extension, modded xorg.conf. 2-finger scrolling now works but tapping is totally disabled. No amount of fiddling with parameters that looked likely, had any effect.

How about trying to use that which was in the previous kernel? I notice from dmesg that previously the input devices were shown a bit differently. I suppose that is due to the new kernel?

[lucky13]

My AA1 has only one card reader (in the right side) that has always worked without problems in Windows and Linux. It is the way I boot normally TC. I have tried with a usb card adapters with the same results. Booting from a usb pendrive has worked in the same way as with the SD card (and I prefer the SD card as it doesn't protrude out of the netbook). And both with fat and ext2 filesystems.

I brought up the card slot(s) because I'm trying to figure out if it's basically like mine has been (this was my first and will be my last Acer). With the udevd error messages you were getting per your comment in the alpha-2 thread I was wondering if the card has been getting caught in some kind of loop or even delay after boot while devices are being populated so that it's not properly reading your tce directory (which reflects some of my own experiences frustrations with my AA1 and why I just use an external reader in Linux rather than deal with the way both readers on my AA1 have proven rather dysfunctional under Linux). But now you write that you get the same errors whether the card is in the internal slot or an external USB reader. I'm trying to figure out how -- barring any corrupt disk or extension(s); please test as curaga suggests -- you'd have no issue under alpha-1 even after trying -2 and -3 unless there's a rule change/conflict causing you some kind of delay since -2 (which is possible but doesn't seem likely from the changelog).

Just to make sure... Did you try alpha-1, -2, and -3 all with both external and internal readers? And did you get similar results either way -- working under -1 internal/external, not working in -2 or -3 internal/external (with comparable waitusb time periods)? And you were able to safely revert to -1 (internally or externally) without any recurrence of the same issue?

[curaga]

Well unfortunately that didn't work so well. Mounted the extension, modded xorg.conf. 2-finger scrolling now works but tapping is totally disabled. No amount of fiddling with parameters that looked likely, had any effect.

How about trying to use that which was in the previous kernel? I notice from dmesg that previously the input devices were shown a bit differently. I suppose that is due to the new kernel?

psmouse.proto should be your option; see kernel-parameters.txt for what protocols it takes.

[jano]

No good news so far:
No errors in the SD card - md5sums of all the extensions are OK.
Reinstalled a fresh alpha-3 in the SD card and checked again for errors and the md5sums -> everything OK.
Booting from the card (directly through the card reader or with a USB reader or USB adapter) with base boot option works OK and I can manually install the extensions without problem, but trying to install only one extension on booting (alsa, pcmanfm or vlc) gives me again booting errors, with waitusb=10.
With alpha-1 (simply overwriting bzImage and tinycore.gz) makes everything work: the three extensions install on boot, without errors, with waitusb=5.

[philip]

Initial setup on my Acer Aspire One netbook model ZG5 is going well. I'm using 32-bit everything, and happy to be submitting this forum post using 3.0a3 with a wireless link. Suggestions:

1. Make at least the wireless extensions easy to find and download without special tools like appbrowser, so users like me can grab them using some other distribution where the wireless works and stuff them into an appropriate tce directory or usb device before booting the alpha release, because its wireless is not working *yet*.

2. Consider putting the "o" back into the "Dwnload" label on the Appbrowser button. (Minimalism is great, but that's one "o" too far.  )

3. Update the info file for package wireless_tools.tcz so it keeps up with the current kernel version. Right now the info file mentions 2.6.29.1, but the dep file cites wireless-KERNEL.tcz.

Warm regards to all contributors!