>>The majority of XP users aren't using that many old apps anyway...
I know Microsoft is a company that typically caters to the masses, but no paying customer should get screwed on backwards compatibility. Microsoft should continue with backwards compatibility for XP, and instead apply their new security model to Longhorn. It's simply too late for them to try to erase old mistakes in existing software.
Ideally, perhaps, if Intel would GPL the whole BIOS, and I'm not talking about the table scraps they CPL'd a few weeks ago. Such routines would have to be completely open for *EVERYONE* (Not just NDA holders like Microsoft).
Making BIOS calls for everything would also make porting code across architectures a nightmare (Even more so).
You hit it more or less in your second sentence. LinuxBIOS is not an OS in a BIOS. It's a minimal BIOS that can bootstrap a system and load a Linux kernel stored in the firmware. This is good because, as you pointed out, there are already tons of drivers and stuff for Linux, and Linux can handle things like initializing hard disks and NICs. It can then load the OS.
LinuxBIOS fits more in the description you gave in your second sentence.
Because often times the SATA controller is not included in the integrated southbridge. Unfortunately the site is slashdotted so I don't know exactly which motherboards we're dealing with, but an example I have off hand is the Tyan S2885 with uses an AMD8111 southbridge and Silicon Image SATA controller seperately. Sort of like how many mainboards with three or four IDE disk connectors or perhaps SCSI support use a Promise, Highpoint, Adaptec, and other PCI devices for the added functionality.
This is stupid. Do people actually think that sentencing one guy to jail will deter the other thousands of spammers, many of whom aren't even subject to the same laws?
Even if it does succeed on a measurable scale, it'll just make those remaining to send spam that much richer.
They don't have an official driver for that combination yet, but you can get an unofficial one that will build on AMD64 with a 2.6 kernel here. I'm currently using it for a Tyan s2885 and it works quite nicely. The performance isn't what I'd expect, but hopefully that will be fixed with the upcoming Detonator/Forceware 6xxx.
I take it the person who asked the question wants to run on server boards, which basically excludes any of the cute little shoebox cases available. They can also be rather large, which excludes many desktop mid-tower cases as well. Most 1U cases I've seen are intended strictly for datacenter use and aren't built with acoustical ergonomics in mind--They usually run several 40mm 7000RPM fans and generate a lot of noise.
I usually just go to Antec for enclosures, but in this case it might just be wiser to have your boss order a few extra nodes that you can take home. Modern Xeon and Opteron systems tend to get very, very hot in 1U enclosures and require wind tunneling for more efficient cooling and require enormous amounts of power from PSUs that cost several hundred dollars in 1U form factor.
At the rate Sun is burning through cash, I wouldn't be surprised if this guy was part of an upcoming round of layoffs anyway. $2 billion is a frickin' ton of money for anyone, but Sun lost nearly half that in their past year alone.
>>I mean does LinuxBIOS initialize hardware and then load my kernel for regular operations?
It initializes the necessary hardware and then loads any "payload." The payload can be anything--A kernel image if you have enough space on your flash part, a bootloader to boot from over a network or local disk, etc.
Currently the Lightning cluster (Ranked at #6 @ top500.org) uses LinuxBIOS to load a bootloader (Etherboot or FILO, I forget which) which then loads a kernel and BProc. No spinning hard disk, no CD-ROM, just an added solid state flash device on the IDE bus containing a bproc-patched kernel with a driver for the interconnect and bpslave. The rest of the operating system for slave nodes, which is only a few kilobytes, is pulled down over the network (Not necessarily ethernet) and loaded into RAM.
>> How are kernel updates done if it is the latter.
Since all the hardware is same for a cluster like Lightning, a kernel update can be done pretty easily with the bpcp (BProc copy) command if you have a new kernel sitting on the master node. I haven't used it in a while, but I think the command goes something like: bpcp vmlinuz 0-N:/boot/vmlinuz, where N is the last node. The syntax is very similar to RSH/SSH, but you can specify as many nodes as you want so you don't need any shell scripts to count up for you. I guess you might also need to tell bpsh to mount a boot partition, depending on how you have it set up.
If you have your kernel in the BIOS, then you have to rebuild and reflash. With the magic of BProc, you don't even need to put the BIOS flashing utility (See their flash_and_burn utility) on the slave node. Just run bpsh 0-N flash_rom newbios.rom. LinuxBIOS even provides fallback functionality so if something goes horribly wrong during this process or the new image doesn't work it can automatically load an old image that does work.
These are just a few of many possible configurations, of course. There are no strict guidelines as to how a LinuxBIOS system must be used. It's extremely flexable, which is one of its main appeals.
Did you RTFA? LinuxBIOS is a BIOS, nothing more, nothing less. They're not trying to put a whole OS on the flash chip, just enough to boot a real OS. In fact, the BIOS is practically designed to be as non-interactive and minimalistic as possible since Linux does in fact take care of much of its own configuration. As Linus commented in setup.S, "We don't need no steenking BIOS anyway (except for the initial loading:-)."
You'll have a hard time finding one pre-loaded with LinuxBIOS, but the VIA EPIA port has come a long way and would make a good starting point. EPIAs are mini-itx boards, usually with VIA EDEN CPUs, that should run for under $500.
Something extremely common like a web search, perhaps? Saying "google" rather than typing in www.google.com might be useful. Or saying "google slashdot" to look for Slashdot on Google without typing. Or programming a voice command to take you to a bookmark.
LinuxBIOS provides fallback capability which allows for a second ROM to be stored on the same chip and be used in case the first one fails. No extra hardware necessary.
Slashdot Mirror
You have to exceed a certain threshold,underclocking will be fine.
I still don't like having Intel telling me what's best for me no matter how I choose to clock my CPU.
>>The majority of XP users aren't using that many old apps anyway...
I know Microsoft is a company that typically caters to the masses, but no paying customer should get screwed on backwards compatibility. Microsoft should continue with backwards compatibility for XP, and instead apply their new security model to Longhorn. It's simply too late for them to try to erase old mistakes in existing software.
LinuxBIOS can load a Linux kernel in the BIOS ROM, but it is not an OS.
>>just wait until the 1st BIOS virus
There have already been several, that was one problem with using DOS.
Ideally, perhaps, if Intel would GPL the whole BIOS, and I'm not talking about the table scraps they CPL'd a few weeks ago. Such routines would have to be completely open for *EVERYONE* (Not just NDA holders like Microsoft).
Making BIOS calls for everything would also make porting code across architectures a nightmare (Even more so).
You hit it more or less in your second sentence. LinuxBIOS is not an OS in a BIOS. It's a minimal BIOS that can bootstrap a system and load a Linux kernel stored in the firmware. This is good because, as you pointed out, there are already tons of drivers and stuff for Linux, and Linux can handle things like initializing hard disks and NICs. It can then load the OS.
LinuxBIOS fits more in the description you gave in your second sentence.
Because often times the SATA controller is not included in the integrated southbridge. Unfortunately the site is slashdotted so I don't know exactly which motherboards we're dealing with, but an example I have off hand is the Tyan S2885 with uses an AMD8111 southbridge and Silicon Image SATA controller seperately. Sort of like how many mainboards with three or four IDE disk connectors or perhaps SCSI support use a Promise, Highpoint, Adaptec, and other PCI devices for the added functionality.
This is stupid. Do people actually think that sentencing one guy to jail will deter the other thousands of spammers, many of whom aren't even subject to the same laws?
Even if it does succeed on a measurable scale, it'll just make those remaining to send spam that much richer.
They don't have an official driver for that combination yet, but you can get an unofficial one that will build on AMD64 with a 2.6 kernel here. I'm currently using it for a Tyan s2885 and it works quite nicely. The performance isn't what I'd expect, but hopefully that will be fixed with the upcoming Detonator/Forceware 6xxx.
Just wait 'till they start displaying ads based on the radio station/CD/MP3/WMV/AAC you happen to be listening to at the time.
I take it the person who asked the question wants to run on server boards, which basically excludes any of the cute little shoebox cases available. They can also be rather large, which excludes many desktop mid-tower cases as well. Most 1U cases I've seen are intended strictly for datacenter use and aren't built with acoustical ergonomics in mind--They usually run several 40mm 7000RPM fans and generate a lot of noise.
I usually just go to Antec for enclosures, but in this case it might just be wiser to have your boss order a few extra nodes that you can take home. Modern Xeon and Opteron systems tend to get very, very hot in 1U enclosures and require wind tunneling for more efficient cooling and require enormous amounts of power from PSUs that cost several hundred dollars in 1U form factor.
Yeah, it would be like Hummer promoting SUVs that have very little lead, but still get 8MPG.
Uhh, is the editor talking about the same company that requires 103W for the latest and greatest processor they have to offer?
Same reason someone would install Linux on an x86 with Windows pre-installed.
That, and OS X is not fully 64-bit yet.
At the rate Sun is burning through cash, I wouldn't be surprised if this guy was part of an upcoming round of layoffs anyway. $2 billion is a frickin' ton of money for anyone, but Sun lost nearly half that in their past year alone.
Because their installers are more newbie friendly.
I've heard a couple variations of this quote that also seem to apply quite nicely. I'll quote/paraphrase the best I can.
It takes a big man to cry, but it takes an even bigger man to eat that man.
It takes a big man to cry, but it takes an even bigger man to laugh at that man.
http://slashdot.org/article.pl?sid=04/03/19/132225 3&mode=thread&tid=137
>>I mean does LinuxBIOS initialize hardware and then load my kernel for regular operations?
It initializes the necessary hardware and then loads any "payload." The payload can be anything--A kernel image if you have enough space on your flash part, a bootloader to boot from over a network or local disk, etc.
Currently the Lightning cluster (Ranked at #6 @ top500.org) uses LinuxBIOS to load a bootloader (Etherboot or FILO, I forget which) which then loads a kernel and BProc. No spinning hard disk, no CD-ROM, just an added solid state flash device on the IDE bus containing a bproc-patched kernel with a driver for the interconnect and bpslave. The rest of the operating system for slave nodes, which is only a few kilobytes, is pulled down over the network (Not necessarily ethernet) and loaded into RAM.
>> How are kernel updates done if it is the latter.
Since all the hardware is same for a cluster like Lightning, a kernel update can be done pretty easily with the bpcp (BProc copy) command if you have a new kernel sitting on the master node. I haven't used it in a while, but I think the command goes something like: bpcp vmlinuz 0-N:/boot/vmlinuz, where N is the last node. The syntax is very similar to RSH/SSH, but you can specify as many nodes as you want so you don't need any shell scripts to count up for you. I guess you might also need to tell bpsh to mount a boot partition, depending on how you have it set up.
If you have your kernel in the BIOS, then you have to rebuild and reflash. With the magic of BProc, you don't even need to put the BIOS flashing utility (See their flash_and_burn utility) on the slave node. Just run bpsh 0-N flash_rom newbios.rom. LinuxBIOS even provides fallback functionality so if something goes horribly wrong during this process or the new image doesn't work it can automatically load an old image that does work.
These are just a few of many possible configurations, of course. There are no strict guidelines as to how a LinuxBIOS system must be used. It's extremely flexable, which is one of its main appeals.
Yes. It's been known to load Windows 2000 in addition to Linux and a couple other *nixes (FreeBSD, Plan9). I'm not certain if WinXp will boot, though.
Did you RTFA? LinuxBIOS is a BIOS, nothing more, nothing less. They're not trying to put a whole OS on the flash chip, just enough to boot a real OS. In fact, the BIOS is practically designed to be as non-interactive and minimalistic as possible since Linux does in fact take care of much of its own configuration. As Linus commented in setup.S, "We don't need no steenking BIOS anyway (except for the initial loading :-)."
LinuxBIOS does that initial loading.
You'll have a hard time finding one pre-loaded with LinuxBIOS, but the VIA EPIA port has come a long way and would make a good starting point. EPIAs are mini-itx boards, usually with VIA EDEN CPUs, that should run for under $500.
It's a BIOS to load an operating system. They've booted Linux, FreeBSD, Plan9, and even Windows 2000.
The trick is finding a bootloader that doesn't depend on legacy BIOS services.
Something extremely common like a web search, perhaps? Saying "google" rather than typing in www.google.com might be useful. Or saying "google slashdot" to look for Slashdot on Google without typing. Or programming a voice command to take you to a bookmark.
LinuxBIOS provides fallback capability which allows for a second ROM to be stored on the same chip and be used in case the first one fails. No extra hardware necessary.