recounting RAM memory when rebooting (Linux is not like Windows)
Can I avoid recounting the RAM memory ("Checking NVRAM") like Windows does when the PC is rebooted? thanks for any help a. (proud using SuSE 8.0, Gnome 1.4, Netscape 7 over K6-II 500mhz, 256ram :) -- Adagilson Batista Bispo da Silva Bibliotecário FIOCRUZ-Centro de Pesquisa Aggeu Magalhães Fone 81 3302-6524
On 01/07/2003 09:59 PM, Adagilson B B da Silva wrote:
Can I avoid recounting the RAM memory ("Checking NVRAM") like Windows does when the PC is rebooted?
This is usually a setting in your BIOS setup. It is independent of your OS
(proud using SuSE 8.0, Gnome 1.4, Netscape 7 over K6-II 500mhz, 256ram :)
Congrats! :-) -- Joe Morris New Tribes Mission Email Address: Joe_Morris@ntm.org Web Address: http://www.mydestiny.net/~joe_morris Registered Linux user 231871 God said, I AM that I AM. I say, by the grace of God, I am what I am.
No, my brother, when I restart Windows it doesn't recount memory because it was done once before while starting up the pc. You're sure, It counts every time I turn on my computer and I know some pcs that doesn't do setting the BIOS setup. But one time it was turned on I can restart windows a thousand times and RAM memory won't be counted again. Is there a way to make Linux handle this the same way? a. Joe Morris (NTM) escreveu:
On 01/07/2003 09:59 PM, Adagilson B B da Silva wrote:
Can I avoid recounting the RAM memory ("Checking NVRAM") like Windows does when the PC is rebooted?
This is usually a setting in your BIOS setup. It is independent of your OS
-- Adagilson Batista Bispo da Silva Bibliotecário FIOCRUZ-Centro de Pesquisa Aggeu Magalhães Fone 81 3302-6524
This has me worried. A few days ago, while yast was installing something with Yast, from my SuSE CD1, it made this LOUD "zip-bang" sound. And yesterday, while making an ISO image of a CD, it made a LOUD "zum"... I know my translation of the noises probably does not help much... the point is, as far as I know my CD works fine, and has always worked, and I never had this problems in windows. And this noises where not SUBTLE, they where pretty loud. Could it be possible that linux is making the CD do things it is not supposed to do (mechanically). I have a Matshita DVD/CDRW UJDA720 if that helps; is in /dev/hdc with SCSI emulation (now /dev/sr0). When I was making the CD ISO image I was using k3b, with generic mmc emulation. I have not changed the original config except for enabling DMA with hdparm (although the first time was with DMA disabled), and entering max_scsi_luns=1 in GRUB, because my DVD/CD was recognized 7 times initially. __________________________________________________ Do you Yahoo!? Yahoo! Mail Plus - Powerful. Affordable. Sign up now. http://mailplus.yahoo.com
On Tuesday 07 January 2003 9:18 am, Marino Fernandez wrote:
This has me worried.
A few days ago, while yast was installing something with Yast, from my SuSE CD1, it made this LOUD "zip-bang" sound. And yesterday, while making an ISO image of a CD, it made a LOUD "zum"...
Just trying to envision what those "sounds" would sound like, and one thought that pops into my mind is the noise that most modern drives make nowadays when an "unbalanced" disk is in the drive [I think the term is "cavetating" -- basically a harmonic vibration as the disk spins] though the "bang" part might be a failed capacitor [which would generally kill the drive outright, so that must not be it...] Is there any way you can force the speed of reading/writing to be lower [less likely to "cavetate..."]
On Tuesday 07 January 2003 7:57 am, Adagilson B B da Silva wrote: [condensed: BIOS checks memory -- not an issue; windows DOESN'T re-check memory, "that's a cool thing"; linux DOES recheck memory, how to avoid?] Pardon me for being silly, but why is this a "problem"? I have a roughly similar system to the one you reported: celeron P-II, 466, 256m of memory; set for "dual boot" -- from the time I select "linux" or "windows" from the grub menu to the time when I can "actually DO something", it seems subjectively faster (shorter?) under linux than it does with windows [and yes, that is with a "pause for logon & validation" that occurs for both linux and windows -- I have another linux system acting as a PDC, so even under windows I have to "log on"] The point here being that the time spent "checking nvram" is so short I hadn't even noticed that it wasn't being done by windows. Since you indicated use of netscape, that would imply using this system as a client rather than as a server [if used as a server, I *could* see justification for making the reboot time as short as possible -- you want the server to be "serving" as soon as possible in the event of an "unplanned" restart...]
On Tue, 2003-01-07 at 09:57, Adagilson B B da Silva wrote:
No, my brother, when I restart Windows it doesn't recount memory because it was done once before while starting up the pc.
win95/98 and 2000 (I am not sure about XP) DO recount the memory during boot. The only difference is that Linux shows you what it does, while windows does it during their splash screens. While I don't know how to turn such a "feature" off, I would suspect that it may cause instability with your O/S by doing so. To my understanding the BIOS check is to see which banks are used, and how much is in each for basic communications, with the option of a very basic RAM test. Whereas, the O/S level RAM check is to check for un/used memory addresses. If the O/S does not have this information prior to loading any info into the addresses, it would open the ability to cause a memory conflict (by trying to load new info into an already used mem address) and thus cause instability, to say the least. This is often the cause of the BSOD in windoze when a piece of software is hard coded to a specific memory address which is already in use by the O/S or another program, or if windows does not properly unload the mem addy, or if windows "forgets" what it put where in the memory. -- JericAtSbcglobalDotNetwork 11:12am up 9 days, 11:59, 10 users, load average: 0.54, 0.29, 0.25
I'm curious what kind of PC this is (Vendor, BIOS, chipset, CPU or MB name). Could it be that when you initially hard boot up your PC it does memory check through the BIOS call (POST). When it's already running and you do soft reboot to Windoze it does not perform memory check because there is no need to. I don't know if Windoze allows to do soft boot to Linux, but Linux does allow soft reboot to any OS. In any way, initial boot up memory check is a feature of BIOS not the OS. You have to check your BIOS setting or documentation. Regards. Alex -------------------
No, my brother, when I restart Windows it doesn't recount memory because it was done once before while starting up the pc. You're sure, It counts every time I turn on my computer and I know
some pcs that doesn't do setting the BIOS setup. But one time it was turned on I can restart windows a thousand times and RAM memory won't be counted again. Is there a way to make Linux handle this the same way? a.
Joe Morris (NTM) escreveu:
On 01/07/2003 09:59 PM, Adagilson B B da Silva wrote:
Can I avoid recounting the RAM memory ("Checking NVRAM") like Windows does when the PC is rebooted?
This is usually a setting in your BIOS setup. It is independent of
your OS
-- Adagilson Batista Bispo da Silva Bibliotecário FIOCRUZ-Centro de Pesquisa Aggeu Magalhães Fone 81 3302-6524
-- Check the headers for your unsubscription address For additional commands send e-mail to suse-linux-e-help@suse.com Also check the archives at http://lists.suse.com Please read the FAQs: suse-linux-e-faq@suse.com
Alex Daniloff wrote:
I'm curious what kind of PC this is (Vendor, BIOS, chipset, CPU or MB name). Could it be that when you initially hard boot up your PC it does memory check through the BIOS call (POST). When it's already running and you do soft reboot to Windoze it does not perform memory check because there is no need to. I don't know if Windoze allows to do soft boot to Linux, but Linux does allow soft reboot to any OS. In any way, initial boot up memory check is a feature of BIOS not the OS. You have to check your BIOS setting or documentation.
Coldboot = the first boot after power-up, it runs every test that had enabled at bios, including extenssive or "fast" memory test. It will be executed too if you, using any method, forces the program to restar at memory address ffff ! In the times of DOS, I'm use a very simple assembly routine to dos this after a defragmentation operation ;- ) Warmboot = In this case you just call the "bootstrap" ( AKA resset ) int and the system starts to boot-up WITHOUT POST !!! I'm not shure, but probably editing lilo/grub you can do it . It is completly independent of BIOS or Motherboard vendor, it is standad even at the first IBM PC, before XT release !!! And YES, rWindows does exactly this when you restarts machine !!! Fábio Rabelo
On Tue, 07 Jan 2003 13:57:24 -0200
Adagilson B B da Silva
No, my brother, when I restart Windows it doesn't recount memory because it was done once before while starting up the pc. You're sure, It counts every time I turn on my computer and I know some pcs that doesn't do setting the BIOS setup. But one time it was turned on I can restart windows a thousand times and RAM memory won't be counted again. Is there a way to make Linux handle this the same way?
You are not talking about rebooting, you are talking about "restarting Windows", which is different. I think I know what you are talking about, when Windows asks "Do you want to restart Windows now?" and then the GUI goes black, and then comes back. Windows never really gives up control of the machine when you do this, it just shuts down far enough that it can reload new drivers, then restarts the GUI. Maybe in linux you can do it by changing from runlevel 5 to runlevel 1, then back to runlevel 5 ? -- use Perl; #powerful programmable prestidigitation
On Tuesday 07 January 2003 10:22 am, zentara wrote:
On Tue, 07 Jan 2003 13:57:24 -0200 Adagilson B B da Silva
wrote: No, my brother, when I restart Windows it doesn't recount memory because it was done once before while starting up the pc. [...] You are not talking about rebooting, you are talking about "restarting Windows", which is different. I think I know what you are talking about, when Windows asks "Do you want to restart Windows now?" and then the GUI goes black, and then comes back.
And to top it off, in order for windows to do this, you have to do "something special" -- hold down the "shift" key when clicking the "OK" button to restart windows [at least, from the "start/shutdown" menu item/dialog -- I've seen this behaviour in a few "installation" type programs but not very often] without the "shift-click", "restarting windows" does a "warmboot" [resets the CPU, but doesn't do some of the POST items, of which the most notable is the "BIOS memory test"]
Windows never really gives up control of the machine when you do this, it just shuts down far enough that it can reload new drivers, then restarts the GUI.
A "close equivalent" for linux would be "log in as a new user" -- this essentially restarts the X server, but doesn't actually reboot the machine. Windows has this also -- labelled as "log off <username>" in the start menu [rarely used in practice, I've found] but functionally the same [ok, slightly worse than restarting X as many "user level" device drivers are included in the "graphical" startup portion of windows -- things that linux relegates to "not part of X"... :) ]
Maybe in linux you can do it by changing from runlevel 5 to runlevel 1, then back to runlevel 5 ?
I still have to ask, however, just how long does it take to "check NVram" on his system and why is this a concern? I don't want to reboot my system "right now" just to find out how long it takes, but shortly I will be rebooting my system [need to do some video editing, and I don't have cinelerra working with my camera yet, so I *have* to use Adobe premiere...] I'll try to keep an eye out for "how long" this step takes during boot...
The 03.01.07 at 13:57, Adagilson B B da Silva wrote:
But one time it was turned on I can restart windows a thousand times and RAM memory won't be counted again.
That happens only when doing a warm reboot. -- Cheers, Carlos Robinson
Sorry for my ignorance, but I don't know what is a *warm reboot*. I know it reboots Windows without counting memory again (it makes my boot time faster), Linux doesn't. Somebody asked why would someone be interested in this silly thing. Well, I just want my OS *BETTER*. Yeap. I want my digital vassal working for me, in my way, without no requirements :) When you make progress, it means you changed for better. This is progress. Don't make me miss Windows, all right :) a. Carlos E. R. escreveu:
The 03.01.07 at 13:57, Adagilson B B da Silva wrote:
But one time it was turned on I can restart windows a thousand times and RAM memory won't be counted again.
That happens only when doing a warm reboot.
-- Adagilson Batista Bispo da Silva Bibliotecário FIOCRUZ-Centro de Pesquisa Aggeu Magalhães Fone 81 3302-6524
On Thursday 09 January 2003 11:14 am, Adagilson B B da Silva wrote:
Sorry for my ignorance, but I don't know what is a *warm reboot*. I know it reboots Windows without counting memory again (it makes my boot time faster), Linux doesn't.
I think there is a deeper underlying "pavlavian response" problem in the works here: why are you "rebooting" your linux system often enough to be worried/concerned about it? After all, the only time you should be "booting" linux is when the power is first applied -- there is no practical reason [outside of certain hardware changes] for a reboot in many cases. FWIW: the other day when I had to "switch to" windows to do some video editing, and then back to linux for real work, I noted the following: power on boot: does POST w/memory count up exit linux/enter windows: POST w/memory count exit windows/enter linux: POST w/OUT memory count in any case: <del> during POST for access to BIOS, then exit (with or without changes): POST w/memory count In other words, if "every time you booted" was to switch from one system to the other, your complaint would be that WINDOWS does the memory count and linux does not... Many BIOS's offer a "fast boot" option which significantly speeds up the boot process by cutting down the actual "POST" checks [most notably, it merely verifies that "memory exists" at particular locations -- fast count, vs. ACTUALLY checking the integrity of memory at those locations -- slow count] The confusing thing is that the wording is "backwards" -- you have to "disable" the option to actually "check" memory properly In any case, usually pressing <esc> during this process immediately "skips" further memory testing -- sure, "it's a manual operation", but when your system boots "unattended", do you really care how long it takes? [you might if you were serving webpages at a VERY popular site, but somehow I doubt this is the case...]
participants (9)
-
Adagilson B B da Silva
-
Alex Daniloff
-
Carlos E. R.
-
Fábio Rabelo
-
Jeric
-
Joe Morris (NTM)
-
Marino Fernandez
-
Tom Emerson
-
zentara