PC of the size/weight which heavy large "mini-note" can have anywhere in Toshiba LibrettoX0 and CASIO Fiva 10X to several years many these days came out at last. it is much quicker than Fiva10X which I am using, and a price is not too high, either and is rather good -- feeling -- coming out -- I think whether there is nothing

Feature

• Small
• It is light. (820g)
• Crusoe 867Mhz
• LCD 1024x768
• HDD 20G (1.8inch)
• Ethernet built-in
• USB 2port
• AC adapter is small.

A x point

• Irregular key arrangement (a key touch is OKed by Libretto X0 system of former years)
• A character is fine (18 dot and 20dot font under daily use).
• The rear with a battery is too thick.

A kinda works point

• A sound of HDD of operation rings in the ear comparatively. Moreover, it is not so quick (about 15 Mb/S).
• A battery does not have so much (being ability about 2 to 3 hours).
• The cooling fan is attached and a sound of operation is jarring. Although it does not almost get hot by favor

Since the main environment was using Kondara 2.1, it installed これを and arrived to jog dial, the Zoom button, the ThumbPhrase button, and ACPI. Moreover, since the hibernation was also made using swsusp, please make it reference.
Since installation was difficult, a note of it is made.

It is the situation roughly checked by Linux.

It seems that the check of operation of Mr. SAITO who looked at the page was carried out by Debian woody. It is said that an installation disk is not caught in the problem of alim 15x3, but has been installed safely.

1. Standard CD-ROM (CD51A)
2. Standard FDD (UFD5)
3. CD-ROM of Kondara 2.1
4. mbm (boot loader) is entered [ Floppy ].

1. Standard FDD (UFD5)
2. mbm (boot loader) is entered [ Floppy ].
3. Another machine which dedicated the distribution of Kondara 2.1

The boot floppy disk for Kondara VAIO-U1 correspondence bootnet.img.bin please utilize by ます's every を置いて If this is used, it is installable in Standard FDD and a network. Please use as dd if=bootnet-u1.img.bin of=/dev/fd0 etc. As for debian (woody), Vine 2.5, and Turno 7, the standard boot FD also seems to be O.K. Redhat 7.3 is the same as that of Kondara.

If a desktop machine etc. is made into a FTP server and the package of the circumference of kernel is replaced, a corresponded kernel can be used from immediately after installation. Moreover, it can install/apply, without depending on WindowsXP at all. Although the contents are slightly old Custom installation set creation memorandum Does it become が参考?

Isn't mbm required, either, if it will be made only Linux, without carrying out WindowsXP and Dual boot?

If it is going to install Kondara 2.1 as it is, it will stop during starting and will not go well. Since the composition of chipset etc. bears a strong resemblance to VAIO C1MRX Mr. itty's page The page of C1 (MRX) relation, such as a challenge (volume on C1MRX), was made reference at の Linux. Seemingly a problem is in the driver of ALI 15X0, and it will lock during access of an IDE device.
This is the greatest problem about installation to U1.

In the present condition Using only FD, since an installer will not start if pci=off, network installation is X.
Although installation is possible from a standard CD-ROM drive, FDD is required, in order to put in mbm, since starting from the loader of XP was not completed well. after all -- FDD and CDROM -- it was both necessity
If the boot from the loader of XP can be performed, it will be O.K. only in CD-ROM, and if a correspondence boot disk can be created well, I will think that network installation is possible only at FDD.

2002/5/24. Above VAIO-U1 correspondence bootnet.img.bin It can install without using XP if done from, FDD, or a network. This detailed procedure is added later.
For the time being, the following is the installation procedure which used FDD and CD-ROM.

TurboLinux7 which suited at hand by chance, and Kondara2.1 which is my main environment were installed. Moreover, Redhat 7.3 was also installed. He plans to try.

Although it does not, debian has not tried. It is still a report of Mr. SAITO.

The installation itself is not those[ with a problem ](ed) as the ALI m15x3 correspondence module is not on strike of Turbo Linux 7. Since it is caught at the time of a reboot, if it rises by pci=off, and alim 15x3 is cut with kernel re-KONFIGU or a patch is applied, a kernel will operate.

Since the IDE driver of ali is not valid similarly, although Vine2.5 is also installable satisfactory, it will be thought that a problem occurs if it employs efficiently. Is it good that it is installable easily?

Is Red Hat software 7.3 the tendency which resembled Kondara fundamentally? Unless it cut extended functions, such as difference Data Resources, about X just for a moment, it has not displayed well.

It is an installation procedure roughly. the procedure which I did -- since it does not remain as it is, a mistake etc. may occur Please give me indication, such as a mistake.

1. It starts from CD-ROM.

   It installed using VAIO standard CD-ROM (CD51/A). It is the prompt of a boot disk in order to avoid 15xalim3 problem in the case of u1, although the I/O Address of ide is specified using standard CD-ROM for a vaio note.

   linux ide2=0x180,0x386 pci=off 

   When carrying out and starting in the state without PCI, it was able to put in somehow.
   If the boot disk corresponding to ALI is made, I will think that network installation will also be possible. Although it is ばっ dust when making the correspondence installation CD etc., there is no time and it has not poked.

2. Party SHONINGU

   At the time of purchase, WindowsXP Home Edition is installed and it has become C:10G and D:10G. Although it is said that the size of C could be changed when Recovery CD-ROM was used, for the time being, D drive was deleted the whole and used as the Linux partition.

   1. hda1 -- Windows XP
   2. hda2 -- /boot -- or -- /
   3. Following swap, /, home, etc.
   
   
3. Installation

   The rest can be installed almost ordinarily. I think that the head of lilo of the partition which installs linux is good.
   Although it is X servers, and it seems that C1MRX does not operate well in 4.1.0 of Kondara, in U1, it operates satisfactory and bitter taste SERARESHON is also effective.

4. A setup of a boot loader

   The free various functions boot loader which has always been indebted since it did not work, although it is said that lilo could be started using the loader of WindowsXP mbm standard [ ] -- it installed from FDD Since XP did not have DOS mode, FD was required for it.

5. Installation of a VAIO-U1 correspondence kernel

   With more than, since it will stop too if PCI of a kernel is employed efficiently, the kernel of correspondence is installed. A fundamental method is the same as the way of the page [ of Mr. above-mentioned itty ] of C1MRX almost. ac patch and the VAIO-U1 correspondence kernel which also corrected the circumference of sonypi are set also at my place.
   Moreover, after installing with solicitude, there is no failure of mkinitrd and lilo.

   Kernel download

   • This server (It is late.) About 70 kb/s
   • Is it called mirror? (High speed.) Many

   Installation procedure

   1. WindowsXP is once started.
   2. C: Create \tmp etc.
   3. A VAIO-U1 correspondence kernel is downloaded there.
   4. Linux is started by pci=off.
   5. A mount point is created as mkdir/mnt/ntfs.
   6. An NTFS partition is mounted as mount-t ntfs-o ro/dev/hda1 / mnt/ntfs.
   7. The kernel downloaded by 3 is installed.

      rpm-ivh kernel-2.4.18-63ky.i586.rpm    rpm-Uvh kernel-headers-2.4.18-63ky.i586.rpm \  (in fact one line)        kernel-source-2.4.18-63ky.i586.rpm kernel-utils-2.4.18-63ky.i586.rpm 

   8. mkinitrd/boot/vmlinuz-2.4.18-63ky.img 2.4.18-63ky is performed.
   9. An entry is added to lilo.conf and it is lilo command execution.
   
   
6. The kernel which performed swsusp correspondence Suspend/Hibernation Since it puts on there, please also make that one reference.

It is O.K. fundamentally above.

The following patches and correction have been performed in the above-mentioned kernel (63ky).

Moreover, the same patch also as a below-mentioned swsusp correspondence kernel has hit.

If a patch finishing kernel is used, HDD can be ordinarily used by UDMA100.
As long as it tried on hdparm-t - From X66 (mode2=UDMA33) - Even if it tries in order to X69 (mode5=UDMA100), about 15 M byte/S and speed hardly change. It seems to reach the ceiling by the performance of HDD.

Especially, satisfactory, it is the thing of Kondara2.1 standard and is operating. Bitter taste SERARESHON is also effective and it is fairly quick.

Postscript: A graphic tip is at XFree86 4.1.0 somehow, although it is ATI Radeon-M. It seems that it is not supported formally. Since, as for Kondara, the correspondence patch is contained It seems that it is satisfactory. Those who are using other distribution are about XFree86 4.2.0. Please try. although it is an object for Kondara too -- the above -- Mr. itty Linux is challenged (PGC-C1MRX). since -- 4.2.0 can be downloaded

Incidentally, also by Vine2.5, it is an attached thing and operated without the problem.

If not only 1024x768 but 800x600,640x480 is specified to be Modes of a Screen section by /etc/X11/XF86Config-4 file Num Lock is carried out. It is on a virtual screen by pushing Ctrl+Alt+'+' or '-'. The change of resolution can be performed.

Since the Zoom button is acquirable if the patch to sonypi is applied, although mentioned later Although it is a virtual screen too in adding a hand to sjog Display expansion / reduction can be performed in one-touch like WindowsXP.

Since it was hopeful XF86Config-4 and XFree86.0.log.u1.txt are exhibited by reference. It is [ whether it becomes helpful and ] ...

• XF86Config-4.u1.txt
• XFree86.0.log.u1.txt

Incidentally XF86Config-4 are close to what Xconfigurator almost made.

Although it is O.K. generally in a standard sound function, cautions are required a little. tone quality ---like -- built-in speaker =x and headphone = -- is it so so?

Although it is a problem

1. A noise rides during reproduction.
2. The volume control of a built-in speaker is PhoneOut.
3. At Alsa, it is useless ? (it has seldom poked).

As a measure

1. It seems that it will generate if processing has not fulfilled demand, although it is a noise under reproduction. By setup of client applications, such as xmms and arts, or a server It seems that it is OK when a buffer is enlarged as much as possible.

   xmms -- [option]-> [a setup] -- buffering -- about 1 second -- if it is alike and carries out A noise is almost lost.
   At KDE, it is a setup of a sound server.

   1. Real-time scheduling is confirmed.
      At this time, a suid bit is stood to /usr/bin/artswrapper. (Unless it carries out this, it does not actually become effective)
   2. A buffer is made into the maximum.

   It comes out and a noise is almost lost.

2. Next, it is although it is the problem of PhoneOut. It will be a PhoneOut=>Mono master if the driver of ac97 is seen. Built-in speaker = it seems that a monophonic recording, a headphone stereo cassette tape recorder, and a master are divided. Therefore, I consider that it is better to carry out volume adjustment by PCM of being main.
   Although it can do in xmms, since there are some which adjust a stereo master, I consider that it is better to adjust PCM from the screen of a mixer etc. by the applet for volume control.

Although sound did not come out in Alsa, it may arrange, when adjusting PhoneOut.
- Although > kernel was reconstructed and retested, it was useless too. Those who used well, if a way can be taught and obtained, it is kind.

It is visible to /dev/sda1. It mounts and reading and writing are possible.

It was suspensible by swsusp. Thank you for Mr. [ Mr. Ogawa who obtained information and report, and / itty ]

To the kernel 2.4.18 of base SourceForce since -- the 2.4.18 correspondence versions of acpi Mr. Florent Chabaud's page since -- swsusp patch dealing with 2.4.18+acpi others -- it is operating safely by kernel which applied patches, such as CLIE ID relation, ALI DMA correspondence, MemoryStick, and sonypi

O Compiled kernel binary

I think that it can use especially regardless of distribution. once big [ except USB ] on our side -- although it is operating satisfactory Even if there is fault etc., since responsibility is not taken, please use it on your responsibility. Incidentally it is operating by Kondara and Redhat7.3.

• kernel-bin-2.4.18-y1.tar.bz2

O Usage

The archive is dedicated with a relative path. It is developed by boot, lib/modules, and usr/src/linux-2.4.18-y1-/include/from a current directory. Be careful of work enough.

1. The following work is done by root authority.
2. An archive is developed from a root directory.
3. initrd is created as "mkinitrd-f/boot/initrd-2.4.18-y1.img 2.4.18-y 1" etc.
4. The link of a header is created as "ln-s/usr/src/linux-2.4.18-y1 / usr/src/linux."
5. The following entries of lilo are created.

   image=/boot/vmlinuz-2.4.18-y1label=linux-y1initrd=/boot/initrd-2.4.18-y1.imgread-onlyroot=/dev/hda5append="acpi=on idebus=50 resume=/dev/hda3"

Most main hardwares are stopped with a script etc. before suspension. It reboots after a return. Although it is O.K. in general, hotplug has not rebooted well yet. (Therefore since USB does not return a manual return (^^))
Moreover, since a clock serves as time at the time of suspend at the time of a return, it is hwclock. -- It is necessary to perform time 合わせ as hctosys etc. Though pending arrangement, it is the script which I am using. It is necessary to make it operate by root authority.

• hiber.sh.txt

since a swap partition is used for an evacuation domain when using swsusp, although it is for a sense -- swap -- about 1.5 times of a real memory -- it is necessary to secure . Be careful.

O To the direction compiled by itself

It is the patch and procedure when creating the above-mentioned binary.

• http://www.kernel.org/pub/linux/kernel/v2.4/linux-2.4.18.tar.bz2 -- kernel main part
• acpi-20020611-2.4.18.diff.gz -- acpi correspondence patch
• http://fchabaud.free.fr/English/Tricks/Laptop/Swsusp/patch-acpi-acpi20020611-swsusp8.gz -- swsusp patch
• kernel-patches-y1.tar.gz -- Other patches

First of all, kernel is developed and a patch is put in patch-p1 in order in it. make mrproper is performed after that.

• dot.config-u1.txt

A を copy is carried out and it renames to .config.

after -- make oldconfig, make dep, and make -- it make modules and installs Since vmlinuz and System.map are made to a root directory, it copies to /boot/vmlinuz-2.4.18-y1 and /boot/System.map-2.4.18-y1 (or movement). The rest performs initrd creation and lilo edit and, finally is.

A device is created first.

mknod /dev/sonypi c 10 250 

Although sonypi is detected at the time of starting Since a dynamic minor number will stick if nothing is carried out Since this is fixed The following lines are added to /etc/modules.conf.

alias char-major-10-250 sonypi  options sonypi minor=250 

They are Zoom and the ThumbPhrase button with the above-mentioned sonypi U1 correspondence patch. An event can be acquired now. The event of a standby button cannot be taken [ whether it will be taken at ACPI, and ].

In handling of Jog or a button sjog and jogdiald were used. after all -- sjog is used. If "-rb" option is not attached at the time of starting of sjog A setup of brightness becomes a light-and-darkness contrary.

I am tampering with and using a little sjog.

• u1 correspondence patch to sjog (Although it is pending arrangement ...)
• sjog-0.5-2.i586.rpm

With this

• The ThumbPhrase button -> mouse button 2 (middle)
• The Zoom button -> screen resolution change
• The Back button -> mouse button 1 (left)

It can use now, carrying out. Since it is weak, pushing a middle button by 2 button simultaneous aggressiveness finds it useful unexpectedly.

Although it is cautions, only the event when being pushed can be taken at sonypi (?)'s. The drug in ThumbPhrase and the Back button is not made (^^;;

/dev/cpu/0/msr is received for changing a setup of LongRun by user authority. Writing needs to be made.
The permission of owner of msr is 600 in root.root. Or [ it operates client applications (gkrellm etc.) by suid of root ] A permission is set to 660 although it is slightly dangerous systematically. If he is put into the root group of /etc/group, I will think that it is good.

1. longrun-0.9 -- Command line tool
2. GKrellMLongRun -- Plug-in of GKrellm

gkacpi は It is the acpi client of gkrellm plug-in. The monitor of a battery, AC adapter state, the temperature, etc. can be carried out. However, it is because of a correspondence ACPI version being old for a while. It will be O.K., if it corrects suitably and installs. Simple and light one is an advantage.

although the effect is not actually measured the floating point arithmetic of crusoe is weak -- since that is right the way of mpg321 (Kondara 2.1 standard and thing same as mpg123) -- mpg123 others -- it is thought that load falls from the thing using a floating decimal point When it is thought by other distribution that load is strangely high How about checking? Since a fixed decimal point can be used also by mpg123 If this is used, I will think that it is good.

Since security update of mpg321 has appeared in 2002/5/19 in Kondara I consider that it is better to be used there.

It is C1 relation and they are a large number. Only a part is ...

1. Mr. Bluce Howard pcg-c1mrx and pcg-c1mv linux configuration
2. Mr. itty Linux is challenged (volume on C1MRX).

I received many advice and kernels directly indirectly especially. Mr. itty Thank you very much.