High amount of caching

Hello, all. First post about arch. Cool distro.
I'm pretty new to Arch but I've been using mainly Fedora for awhile now. I've noticed something odd in Arch on my most recently acquired box.
Pentium 4 3.0ghz
2x 80GB HDD
1GB of memory
ATI Radeon AGP card
Arch 2008.6
I've noticed unusually high amounts of memory caching - unusual at least in my experience. The true usage is very low but I've been noticing the kernel caching about 800-900MB at any given time. The box is currently running without X as I try to troubleshoot my xorg.conf (that's an issue for another post). top, free -m, and /proc/meminfo routinely point to high amounts of memory being cached but actual usage has been hovering around 100MB at most.
$ free -m
total used free shared buffers cached
Mem: 1010 995 15 0 49 885
-/+ buffers/cache: 60 949
Swap: 1929 0 1929
$ cat /proc/meminfo
MemTotal: 1034704 kB
MemFree: 15400 kB
Buffers: 50436 kB
Cached: 906848 kB
SwapCached: 0 kB
Active: 40992 kB
Inactive: 921604 kB
HighTotal: 130512 kB
HighFree: 876 kB
LowTotal: 904192 kB
LowFree: 14524 kB
SwapTotal: 1975912 kB
SwapFree: 1975912 kB
[snipped]
$ top
top - 15:32:49 up 18:04, 1 user, load average: 0.00, 0.00, 0.00
Tasks: 64 total, 1 running, 63 sleeping, 0 stopped, 0 zombie
Cpu(s): 0.1%us, 0.2%sy, 0.0%ni, 99.3%id, 0.3%wa, 0.0%hi, 0.1%si, 0.0%st
Mem: 1034704k total, 1019416k used, 15288k free, 50536k buffers
Swap: 1975912k total, 0k used, 1975912k free, 906848k cached
PID USER PR NI VIRT RES SHR S %CPU %MEM TIME+ COMMAND
1 root 20 0 1664 560 492 S 0 0.1 0:01.09 init
2 root 15 -5 0 0 0 S 0 0.0 0:00.00 kthreadd
3 root RT -5 0 0 0 S 0 0.0 0:00.01 migration/0
4 root 15 -5 0 0 0 S 0 0.0 0:04.53 ksoftirqd/0
5 root RT -5 0 0 0 S 0 0.0 0:00.00 watchdog/0
[snipped]
However, the machine is very responsive and is handling Samba service very quickly. I should mention I have one 1 GB swap per hard drive for a total of 2GB swap, but I've not even seen those swaps in use yet. Do I need to be concerned or is the high amount of caching something I should look into further? I know it's common for the kernel to cache some, but I've never seen the kernel cache ~95% of the system's available memory.
Thanks. Let me know if you need more output or further specs.

switch86 wrote:
fwojciec wrote:Would you rather have your system read everything from the HD over and over again?
No, definitely not. I realize that caching is not the same as "using," but it seemed like a high amount of caching. I'm being paranoid, then?
Yes, you're. But the term is "ricer" I guess
Seriously now... you can argue what is a high amount of cache with kernel developers, they know better.

Similar Messages

Invoice Reduction-Vendor error: only a higher amount invoiced is allowed

Hi,
I am trying to replicate the "Invoice Reduction" Scenario (for both Price and Quantity Variance scenario).
1) Created a PO for 25 PCE and made GR for 10 PCE (10 USD/PCE).
2) Entering Invoice for 25 PCE and Amount 2500 USD and I have selected correction ID as Vendor Error: Reduce Invoice. So when the document is posted system should generates the Invoice document and Credit memo.
But I am getting the error as mentioned in subject.
Please find below the Error description:
Vendor error: only a higher amount invoiced is allowed
Message no. M8582
Diagnosis
The item amount in the invoice of 100.00 is smaller than the item amount of 2,500.00 suggested by the system.
System response
You currently cannot enter an amount that is smaller than that expected by the system. This would lead to a credit memo for the vendor.
Procedure
You can change the default values to correspond to the invoice and then post the invoice.
Regards,
Hari.

Hi,
Please check if the PO is GR based invoice verification is checked.
since the GR is only done for the partial qty while you enter the invoice and reducing it system is giving the error.
If vendor sends a higher amount /qty when compared to PO you can try for a reduction.
for more information on when and how an invoice reduction need to be done,please check the below link.
http://help.sap.com/erp2005_ehp_04/helpdata/EN/a8/b99284452b11d189430000e829fbbd/frameset.htm
and http://help.sap.com/erp2005_ehp_04/helpdata/EN/d6/36451bca1a4a81b974bb9114e41458/frameset.htm
and
http://help.sap.com/erp2005_ehp_04/helpdata/EN/a8/b9975e452b11d189430000e829fbbd/frameset.htm
Regards,

Warning Pop Up message while inventory adjustments with high amounts a

Dear Friends,
I had a issue.
I would like to let appear a warning popup window when inventory adjustments with high amounts are entered. This warning should be accepted before posting movement types when the amount is exceeding 10,000 USD.
Please give solution.
Regards,
Mahesh

Hi,
Its possible through BADI: Ex:MB_MIGO_ITEM_BAdI.
but for your suitable BADI , you have search at
spro -mm- inv.mgt & phy.Inv-Maintain customer exits & Business Addins.
regards

To post invoice with higher amount

Hi all,
We want to try out a scenario, where in we want to post invoice with higher amount than PO. This is owing to big lead time between ordering and delivery. (And this is not the case of currency exchange rate changes). Can anyone suggest a solution / FI settings to do Invoice posting.

hi,
I see two ways:
1. You can see the setting for the over posting in the MIRO itself...FOr this go through:
Spro --> MM --> Logistics invoice verification --> define Vendor specific tolerances --> Here you can provide the tolerances as per your requirement...Here check for Automatic accpetence of the positive differences....
2. You can use the tolerance B1 for the GR where you can post for the more amount , only you get the mail for more amount posted, and use the GR based IR indicator in the po...so taht invoice will be made of the same amount...
Check the path for the tolerance setting:
Spro --> MM --> Inventory mana. and PI ---> GR --> Set tolerance limit --> here enter for new details...
Regards
Priyanka.P

Igsmux_mt consumes high amount of CPU

Hi experts
IGS is configured on 2 ports 40101 and 40102 hence 2 igspw_mt process
now there are 2 igsmux_mt process
1 is consuming around 60% of CPU
Name            PID CPU% PgSp Owner           ServerV3       0   "h" for help
igsmux_m     557252 61.5 11.9 ap5adm          ClientV3       0   "q" to quit
java         446516   8.9 59.7 root
plaixlps161:ap5adm 6> ps -ef|grep -i igsmux
ap5adm 557252 806988 480   Sep 12      - 14461:03 /usr/sap/AP5/SYS/exe/run/igsmux_mt -mode=profile -restartcount=0 pf=/usr/sap/AP5/SYS/profile/AP5_D01_plsapap501
ap5adm 590052 1306756   0 09:43:37 pts/1 0:00 grep -i igsmux
ap5adm 610450 393276   0   Sep 12      - 2:45 /usr/sap/AP5/SYS/exe/run/igsmux_mt -mode=profile -restartcount=0 pf=/usr/sap/AP5/SYS/profile/AP5_D01_plsapap501
As there are 2 port watcher process der r 2 igsmux_mt process rt so if i kill 1 the high consuming process it will be restarted
is it ok if i kill it?
any idea to troubleshoot why this is consuming high amount of CPU?
Also when i test connection it shows IGS could not be reached
Awaiting reply
Regards
Ameya
Edited by: Ameya Joshi on Sep 16, 2009 10:03 AM
Edited by: Ameya Joshi on Sep 16, 2009 10:09 AM

Thanks Juan
I already read that thread
In it is mentioned igswd_mt i.e watchdog thread automatically restarts the port watcher process igspw_mt rt
Also 1 more thing
in transaction SIGS , in the adiministration menu there are no incoming calls i mean the calls graph is at 0
now the mux process listens the call from listener and forwards it to port watcher rt
now as no calls are the mux should not be active rt?
Also when i test the chart engine it shows IGS not reached
Awaiting your reply

IPhone using extremely high amounts of data when not in use

I have a 4S on AT&T and am grandfathered in w/ the unlimited data plan. Last month I got the notification that I was in the top 5% of data users (which sounds crazy to me considering what some other people I've read about seem to use). According to AT&T's site I had used 2.09 GB with 16 days left in my billing cycle.
Once I got that notification I logged on to see if I could find out when I was using such large amounts of data. It turns out between midnight and 1 am almost every night there was large amounts of data sent. I think this is strange since for two reasons, 1) I'm on WiFi at home in my apartment and 2) I wasn't even awake at the times these charges occurred.
Example: January 6, 2012 at 12:40 am I was charged for 417,327kb, roughly 417MB.
Similar data usage occurred almost every night going back to 12/28/11. On 12/29 I was charged for 468MB at 12:21am. Very unlikely this was actually data I used since I had work the next day and wasn't awake. It honestly looks like 80-85% of my data usage is coming from these occurrences. It also isn't likely that this is a total of data used throughout the day as there are other entries of smaller amounts spread out throughout the day.
Now, if I was on a truly unlimited plan and there was no such thing as throttling I really wouldn't care about this. But the fact is that my 3G speeds are being throttled (just ran speed test at a location I used to get over 1Mbps and I am was at .07 Mbps). I have spoken to AT&T and they insisted it was an issue w/ my phone, either the hardware or the OS. So I went to apple and the "Genius" did a DFU restore for me in store and told me that would fix it. It hasn't and now I'm stuck with this constantly happening and unbearably slow 3G data speeds for the rest of the month.
That was my last billing cycle where I was throttled down to download speeds of .07mbps. Unusable. This month I decided to closely monitor my data usage. I turned off iCloud and photostream, and turn off data whenever I'm using wifi. I turned off the sending info to apple setting.
I checked my account on ATT.com today and noticed that yesterday morning at 7:58 am while driving to work I somehow used 247 MB. Don't ask me how, I'm not steaming anything and my phone is either locked or playing music from the iPod app. I could probably stream Netflix for 5 hours and not use that much data.
I called AT&T to let them know (AGAIN) and got bumped up to a "Data manager" who was incredibly b*tchy and rude. She said there must be some app that's using that much data to update. I don't have any apps that are even that size so I don't see how this could possibly be the case. I tried to talk with her about how that can't really be the case and she just kept repeating that there must be some app using the data blah blah. I've gone to apple with this last month after failing with AT&T and at first they did a DFU restore and we restored the phone as new, then it kept happening so I went back they replaced my phone. AT&T said it must be a hardware or software issue last month, but this month it must be an app that's doing this. I can't get a straight answer from them and they just keep passing the ball to say it's Apple's problem.
I really don't know what to do. I'm grandfathered in on the unlimited plan and don't really want to change that. Say I change my plan to the 3GB's at the same $30/month. What's to say this won't keep happening and I'll be over that 3GB's in 10 days then get charged an extra $10 for each additional GB?!
I can't think of any apps I've downloaded in the past few months that would result in this change and I can't live with another 2/3 of my billing cycle being throttled down to unusable speeds.
Has anybody else had anything similar happen? I have no idea what to do next (and sorry for the long rant but I'm fresh off the phone with AT&T and I'm ******)

Thanks for the link. I only went through the last three pages and unfortunately it looks like there is no solution. I turn data off every night and turn it on only when I'm not in a wifi area. It seems that as soon as I turn it on, within an hour it charges me data for a backlog of whatever it didn't do when it was connected to wifi.
I think it's absolutely disgusting that AT&T can just dismiss this and act like it's the users fault. Apps and email updating in the middle of the night to the tune of 400MB? I highly doubt it and then when I called to ask them about it they try to make me sound stupid like I have an app or two open. Just really ticked me off. The worst part about it is that there doesn't appear to be anything I can do about it aside from never using my data.

Cpu usage high when loading cache group

Hi,
What are the possible reasons that results high cpu usage when loading read-only cache group with big root table (~ 1 million records)? I have tried setting Logging=0 (without cache agent), 1 or 2 but it doesn't help. Are there any other tuning configuration required to avoid high cpu consumption?
ttVersion: TimesTen Release 6.0.2 (32 bit Solaris)
Any help would be highly appreciated. Thanks in advance.

High CPU usage is not necessarily a problem as long as the CPU is being used to do useful work. In that case high CPU usage shows that things are being processed taling maximum advantage of available CPU power. The single most common mistake is to not properly size the primary key hash index in TimesTen. Whenever you create a table with a PK in TimesTen (whetehr it is part of cache group or just a standalone table) myou must always specify the size of the PK hash index using the UNIQUE HASH ON (pk colukns) PAGES = n clause (see the documentation). n should be set to the maximum number of rows expected in the table / 256. The default is sized for a table of just 4000 rows! If you try and load 1M rows into this table we will be wasting a lot of CPU time serially scanning the (very long) hash chains in each bucket for every row inserted...

High Amount of Spam on Exchange 2013 - Content Filtering is Enabled but Pfizer Spam Filling Up Everyone's Mailboxes

Hello
Previously I used Exchange 2010 with Forefront Threat Protection installed and this used to do a good job of stopping all the spam.
However since updating to Exchange 2013 earlier this year and enabling the integrated spam filtering everyone noticed a sudden increase in the amount of spam which was getting through which has been bad for a long time.
We have been living with it but in the last 3 weeks everyone has started getting about 40 emails a day from Pfizer for Viagra. All these seem to defeat the content filtering as Viagra is spelt with an extra I and the email address is always different.
Also images in emails are blocked by default but somehow all the images on these spam messages appear for everyone.
I am not sure the spam filtering is working at all and I'm not sure how to tell as ForeFront gives you a nice graphical dashboard but I can find nothing similar to this in Exchange and PowerShell seems the only way to configure the limited functionality
of the content filter.
Is there any way to get rid of these messages as it doesn't look very good when they are constantly popping up for everyone?
Thanks
Robin
Robin Wilson

Hello ManU
Thanks for the reply.
I have checked the logs and see this quite often:
AcceptMessage,,SCL,not available: policy is disabled
But other times it says this:
RejectMessage,550 5.7.1 Message rejected as spam by Content Filtering
Which seems to indicate it is rejecting some.
This is what one of the email headers look like:
Received: from RWS-MAIL.rwsservices.net (192.168.2.151) by
RWS-MAIL.rwsservices.net (192.168.2.151) with Microsoft SMTP Server (TLS) id
15.0.775.38 via Mailbox Transport; Sat, 28 Dec 2013 10:59:26 +0000
Received: from RWS-MAIL.rwsservices.net (192.168.2.151) by
rws-mail.rwsservices.net (192.168.2.151) with Microsoft SMTP Server (TLS) id
15.0.775.38; Sat, 28 Dec 2013 10:58:38 +0000
Received: from [90.169.106.204] (90.169.106.204) by mail.rwsservices.net
(192.168.2.151) with Microsoft SMTP Server id 15.0.775.38 via Frontend
Transport; Sat, 28 Dec 2013 10:58:37 +0000
Date: Sat, 28 Dec 2013 12:05:58 +0200
From: US.Pfizer eStore <[email protected]>
To: robin.wilson <[email protected]>
Message-ID: <[email protected]>
Subject: Dear robin.wilson up to 65% OFF!
X-Mailer: Airmail (223)
MIME-Version: 1.0
Content-Type: multipart/mixed; boundary="dd2ee3ea_586bb9e4_6f04"
Return-Path: [email protected]
X-MS-Exchange-Organization-PRD: 001-taxis.co.uk
X-MS-Exchange-Organization-SenderIdResult: Neutral
Received-SPF: Neutral (rws-mail.rwsservices.net: 90.169.106.204 is neither
permitted nor denied by domain of [email protected])
X-MS-Exchange-Organization-Network-Message-Id: e8825204-1f32-48be-a331-08d0d1d30209
X-MS-Exchange-Organization-SCL: 1
X-MS-Exchange-Organization-PCL: 2
X-MS-Exchange-Organization-Antispam-Report: DV:3.3.13223.464;SID:SenderIDStatus Neutral;OrigIP:90.169.106.204
X-EXCLAIMER-MD-CONFIG: 079171ba-394f-46d5-a160-56e416712e8e
X-MS-Exchange-Organization-AVStamp-Enterprise: 1.0
X-MS-Exchange-Organization-AuthSource: rws-mail.rwsservices.net
X-MS-Exchange-Organization-AuthAs: Anonymous
The emails use a different sender email address every time and there is always a poem in very light grey writing in the body of the email. The drugs are always misspelt as well. Is this why these are getting through?
Thanks
Robin
Robin Wilson

Thunderbird crashes when importing high amount of Outlook Express .dbx files.

Greetings !
I'm trying to make a migration from Outlook Express to Thunderbird for a professional. His mail folder is made of around 1400 .dbx files (18GB of mails -woh-). Every time I'm trying to do the importation, Thunderbird crashes.
I think I should specify that this importation is happening on the new computer I set up and from the old hard drive.
I also tried importing small amounts : it worked but it's killing the file tree, which is not acceptable on a such amount of mails.
I found mention of an 'Undbx' software that could help me. I'll try it very soon and report.
I will provide a dxdiag.txt soon too, but I really think the PC is set up correctly (I have some experience in this domain).
Importing this mail folder is critical since it's a professional need, so I'm relying on you and deeply thank you for your effort here.
PS : English is not my native tongue, excuse the grammar and spelling mistakes.

Dear Matt.
I can proudly announced the problem is solved !
In 'accounts settings' (not sure of my translation here), I set the mail storage folder to a new one, near the hard drive root, then launched the importation. And it went just well, keeping up the file tree and all.
Thank you a lot for the informations you gave me. I wouldn't have found out the path length is limited to 255 without you. '''You rock, sir !'''
I owe you.
Keep up the good work. Thanks again.
Sincerly yours,
Jack
PS : Some information on this issue on the database would be welcome I think and I'd happily participate.

Time Machine backing up high amount of GB's after I upgraded to Mountain Lion. Is there a way to correct this?

After upgrading my IMac to Mountain Lion, Time Machine keeps backing up large amounts of data (50 to 60 GB at at time}. I have not imputed that kind of data into the Mac since I have lttle use for it for the moment. My back up disc is a Time Capsule. Help!

OK, I will hold for a couple of days see what happens. First back up after the upgrade was 10 gb. the second one is in progress and is backink up 59.8 gb.
No data inputs have occured since the last back up before the upgrade and usually I only do small tasks such as family photos, email and ocassionaly on line shopping such as a few songs from ITunes.
Thanks I'll keep you informed.

Is there a noticable difference between2.3 ghz, 2,6ghz and 2.7ghz. and is it worth paying to upgrade from 2.3ghz to a higher amount of ghz

is there a noticable difference between 2.3ghz, 2.6ghz and 2.7ghz and is it worth the upgrade to a higher amout of ghz

going from 2.3GHz to 2.7GHz you might notice the difference.....
but going from 2.6GHz to 2.7GHz, don't even waste your money....
if you'll be using it for everyday use, then 2.3GHz will do just fine....
however, if there's photoshop and some video editing happening, then 2.6GHz would be better suited for you....
or, go with the 2.3GHz base model then upgrade it to 16GB of RAM and add a SSD to get the best price - performance ratio...
good luck...

Matching invoice to Purchase Order with higher amount without warning

Oracle Applications : 11.5.10.2
PO was created via iProcurement with line type, service amount(The Value Basis is Amount). The order quantity is 100, unit price is 5, amount is 500
I created a invoice in AP and try to match it to this PO from invoice workbench.
In the Match to Purchase order Window, I changed the quanitity invoiced from 100 to 200, Amount updated to 1000 automatically. Then click the match button. There is a warning: This match will result in an overbill.
But when I update the Match Amount from 500 to 1000, the unit price updated from 5 to 10. And system did NOT give me a warning notice.
I know that Payables automatically calculates and enters the remaining field (Quantity Invoiced x Unit Price = Match Amount).
but can someone tell me why system doesnot give a warning when I update the Match Amount? Is there any setup we can set to let system pop up warning?
Thanks
Kyla

Thank you Gordon,
However I cannot deny to the user that wants to apply that that there is a business case where the vendor demands the invoice to be paid all in front (maybe one installment with the Purchase Order and one with the Receipt of the Goods). So, I would expect this to be covered.
I noticed another thing: When I opened the Down Payments screen on Invoice (where Down Payments are listed in order to choose them) there is a column "Open Net Amount" which brings the correct amount. However, in the column "Net Amount to Draw" there is a wrong figure and that makes the mistake.
Now I can see that I can wright the correct amount to column "Net Amount to Draw" but still I cannot understand why a user should do that.

My Mac is showing a very high amount of used storage for photos

Hi guys,
I'm starting to run low on disc space and decided to see how much of my hard drive is being used for what by using "About This Mac". My Macbook Pro is showing that I have used 164.62GB on photos. I've checked and my iPhoto library (which stores all my personal pictures) is only 68.66GB. I get that there will obviously be more spaced used on my HDD for pictures other than my iPhoto but I really can't see how that totals nearly 100GB. Is there likely to be an old iPhoto library file lurking in a folder somewhere?
I'm pretty sure I'm also not using as much space for movies as well but this is nearly no where as much of a concern.
Any help on what could be causing this would be very much appreciated.
Thanks
MegaBeaker

my hard drive is being used for what by using "About This Mac". My Macbook Pro is showing that I have used 164.62GB on photos. I've checked and my iPhoto library (which stores all my personal pictures) is only 68.66GB.
About this Mac is a very crude measure of anything. That will include just about every image, icon, UI element on your Mac. It's so vague as to be meaningless.
How big is your Pictures Folder?

High amount of daily d/c's and line noise regularl...

Afternoon all, heres my problem/gripe:
I'm a regular online gamer, and i'm finding that i get anything from 8-15 disconnections from the network daily, which, as you can understand, is a lil bit frustrating. I have tried different modem/routers, and, at the moment, the HomeHub 3, although it gives slightly lower speeds than the Netgear DG834v4 I was using before, it's not bad for handling the d/c's, ports are forwarded etc for games....
    Also, i have a lot of line noise on the phoneline using the quiet test, heres my stats below. I find that if I completely unplug the filter, it settles the noise for an hour or so, then its all cruddy again. Im using the filters provided by BT. Also, the speed drops randomly, its pretty stable at the moment, but sometimes can drop as low as 0.98kbps, sending me into the dark ages, having me scrabbling around for my club and loincloth, which is annoying, as when i first moved into this house, the net was a respectable 1.4mbps, but has dropped in about three noticable stages over the 6 months since i've lived here.
     I have tried both microfilters that BT have given me, and, am unable to really get a stable line thru the test socket as it would involve metres of extension cable, which would be a pointless exercise..Oh..and the other annoyance is that every time, without fail, the landline rings...I get a disconnection.
    Apologies for the wall of text..but it's not isolated incidents, but daily, regularly..often.. and ofc, i'm a bit wary of phoning the helpdesk, as, ofc, I don't want to be taught to suck lemons, and want to eliminate any hardare error before risking the £130 callout..
   Apologies for wall of text, I have a VERY manly chest and have a lot to get off it!!!
ADSL Line Status
Connection information
Line state:
Connected
Connection time:
0 day, 00:16:16
Downstream:
9,619 Kbps
Upstream:
1,103 Kbps
  ADSL settings
VPI/VCI:
0/38
Type:
PPPoA
Modulation:
G.992.5 Annex A
Latency type:
Fast
Noise margin (Down/Up):
3.0 dB / 6.3 dB
Line attenuation (Down/Up):
33.5 dB / 18.2 dB
Output power (Down/Up):
0.0 dBm / 12.3 dBm
FEC Events (Down/Up):
0 / 0
CRC Events (Down/Up):
41 / 0

As you can hear line noise that needs to be resolved first If you connect via the test socket and try the quiet line test socket and can still hear the noise then you need to report it to BT faults on 151 but do not mention broadband when you call as you are disconnecting when the phone rings that also implies a line problem providing you connect at the test socket and the problem is still there you will not be charged for an engineers visit
If you want to say thanks for a helpful answer,please click on the Ratings star on the left-hand side If the reply answers your question then please mark as ’Mark as Accepted Solution’

[solved] Booting a "Multiboot" OS only works with high amounts of mem

edit: ok, I hate when it happens, but I've just found the solution: raising the memory reserved for the virtual machine. It's funny, however, that such simple kernel won't boot with less than 128MB of RAM.
I'm not sure if this is the right place to post this question, but as it's at least indirectly related to programming, I'll post it here.
Recently I've read Tenenbaum's book about Minix and now I'm feeling like hacking a simple OS if only for saying "I can do it". The problem is I got stuck right at the first step: booting. I wasn't willing to implement a bootloader so I decided to use Grub and implement a Multiboot compatible kernel.
Running 'info multiboot' you can see a example showing how to implement a "multiboot" kernel. I used exactly the same code and compiled it with:
gcc -nostartfiles -nodefaultlibs -o kernel boot.S kernel.c
The code compiles with two warnings about redefinitions of printf, but it gives no errors. The 'kernel' file is less than 3KB big and running 'objdump -d kernel' seems to be fine.
Satisfied with it, I've tried to create an ISO image so I could test it in a virtual machine (VirtualBox). I followed the instructions from Grub manual of how to install it in a CD (create a iso/boot/grub/ folder, put stage2_eltorito there, etc). I also created a menu.lst like this:
default 0
timeout 30
fallback 1
title Test
root (cd)
kernel /boot/kernel
and created the ISO image with: 'mkisofs -R -b boot/grub/stage2_eltorito -no-emul-boot -boot-load-size 4 -boot-info-table -o grub.iso iso' .
I've tried to run it in a virtual machine and it boots, but when I press enter to run 'Test', which should run the kernel from the 'multiboot' example, grub gives me an error:
Booting 'Test'
root (cd)
Filesystem type is iso9660, using whole disk.
kernel /boot/kernel
[Multiboot-elf, <0x8048000:0x803:0x0>
Error 28: Selected item cannot fit into memory.
Press any key to continue...
And then it fallsback to the menu. (The menu shows [639k lower / 64448k upper memory])
Last edited by andre.ramaciotti (2009-10-12 19:54:04)

The linker is compiling your binary to the default location of 0x8000000. This is fine for Linux to load as an application (even with less than 128MB memory, Linux can use the MMU on the CPU to map anything to 0x8000000). However with ring0 code you are dealing with real physical memory locations. You would need either a linker script (which you can look up, also see: http://wiki.osdev.org/Linker_Scripts) or to specify the options on the command line with
ld -Ttext 0x100000 etc.
0x100000 is at 1MB, so that leaves BIOS memory unaffected. For memory management past that you should read a memory map, which is beyond the scope of this.
Edit: Fix formatting issues.
Last edited by sdkmvx (2009-10-12 19:41:52)

High amount of caching

Similar Messages

Maybe you are looking for