Slashdot Mirror
Linus Says Pre-2.6 is Coming
gomoX writes "As seen on C|Net , Linus has announced that the pre-2.6 series will be starting in early July. Despite not having been able to meet the release goal for 2.6 in June 2003, the next stable version is not as far away as you may think. You can take your guess based on the fact there was a 9 month period between first test version of 2.4 and the official release of 2.4.0 on January 2001."
My question is this:
There was some hesitancy, upon the official release of kernel 2.4, based upon some bugs etc...
I'm wondering, does the kernel - generally speaking - get more and more stable. For example, will the first release of 2.6 be more stable than the first release of 2.4. I realise that there are new additions to the kernel, and with that new problems will probably emerge. However, comparatively speaking, does it make sense that the kernel's evolution will lean towards stability with each release in the cycle, or will it generally be unnoticable?
Just curious.
This wasn't just plain terrible, this was fancy terrible. This was terrible with raisins in it. - Dorothy Parker
I'm sure I'm not the only one who's wondering if Reiser 4 will go into the stock 2.6! So: does anyone know?
"What about that? Will we be finaly able to switch kernels without a reboot?"
I did that back in the 2.2 days with monte. Later with 2.4 kernels I did a few changes, added a feature I was missing, fixed a bug and such stuff. In case you want to see it. But it was never completely stable and lacked SMP support.
kexec might be a better alternative. AFAIK it is being maintained and might even have made it into the 2.5 kernel.
It was only a couple of years ago that knowledgeable people were calling this idea ridiculous, and giving good reasons, however progress has marched on, and we're actually coming within sight of it. The basic challenges are much the same as for hotplug cpus, hotplug memory, process migration in a cluster, and yes, kexec, all of which are being worked on or already working. So I'll go out on a limb and predict that hot-kernel swapping will be demonstrated during the 2.7 timeframe. It won't be perfect, but such things never are in the first cut.
The thing that makes hot kernel swapping practical is the stable api between userland and the kernel. Big changes there are few and far between, and they can be special-cased.
Have you got your LWN subscription yet?
I've been hearing though other channels that the IDE layer rewrite improves the IDE subsystem to the point where SCSI emulation won't be needed to drive an IDE CD burner. Can anyone confirm or deny this? If so, this will probably become my main reason to switch to 2.6 (although there are quite a few secondary ones too). Thanks linux team (and IDE rewrite folks)!
"Can't you see that everyone is buying station wagons?"
You are correct in assuming you don't need ide-scsi to emulate a SCSI host for burning cdroms in 2.6, but it has nothing at all to do with the IDE rewrite.
2.6 has support for queueing "generic scsi" commands through the block layer, using the same mechanism and transport as the regular read/write file system requests. So we can overload the sg (scsi generic) SG_IO and provide the same functionality for non-scsi attached devices (such as atapi burners). With a recent cdrecord, you can give the device with -dev=/dev/hdc for instance.
Additionally, cd burning is now zero copy. The user space data buffer is mapped directly into the kernel for the dma operations. DMA is supported on a 4-byte boundary, where 2.4 and previous has required sector alignment (512 bytes) for any atapi dma operations.