Linux 4.14 Has Been Released (kernelnewbies.org) 89
diegocg quotes Kernel Newbies: Linux 4.11 has been released. This release adds support for bigger memory limits in x86 hardware (128PiB of virtual address space, 4PiB of physical address space); support for AMD Secure Memory Encryption; a new unwinder that provides better kernel traces and a smaller kernel size; support for the zstd compression algorithm has been added to Btrfs and Squashfs; support for zero-copy of data from user memory to sockets; support for Heterogeneous Memory Management that will be needed in future GPUs; better cpufreq behaviour in some corner cases; faster TBL flushing by using the PCID instruction; asynchronous non-blocking buffered reads; and many new drivers and other improvements.
Phoronix has more on the changes in Linux 4.14 -- and notes that its codename is still "Fearless Coyote."
Phoronix has more on the changes in Linux 4.14 -- and notes that its codename is still "Fearless Coyote."
Which is it? (Score:5, Informative)
4.14 or 4.11?
(I expect the summary will eventually get fixed, followed by someone replying to me “WTF are you talking about?”)
Re: (Score:3, Insightful)
I expect the summary will eventually get fixed
A summary actually getting fixed? WTF are you talking about?
Re: (Score:2)
Well, the title says "Linux 4.14 Has Been Released" while the summary begins with "Linux 4.11 has been released."
Welcome to bizarro world!
Re: (Score:1)
> uname -rv
4.13.11-200.fc26.x86_64 #1 SMP Thu Nov 2 18:28:35 UTC 2017
It must be noted that I am running the stable version of Fedora 26 not the developer's version, however, I do have a tenancy to get a new incremental release of the kernel once a week as part of the normal update process.
Of course like most Linux distribution updates I have the choice of a graphical update or command line update or a combination and except for initializing the update process (I co
Re:Which is it? (Score:5, Funny)
It's 4.11 for Workgroups.
Re: (Score:1)
Re: (Score:2)
Re: (Score:1)
Holy inadequate Batman!
Re: (Score:2)
Is there another OS with USB drivers not loaded into kernel space?
Re: (Score:3)
Re: (Score:2)
Windows 7 did, BSOD's were pretty annoying when using Prolific USB serial adapters - the dodgy 3rd party code Windows Update automatically installs and runs in the kernel (like every 3rd party driver) when you plug the USB device in.
Re: (Score:2)
Oddly I can still remember first launching USB under SuSE 9.3 and having to troll my
I probably couldn't even remember how to do that these days.
Re: (Score:2)
I got into a situation last week doing a fresh install where the chipset's USB host support was built as a module but not included in initramfs. A startup problem (fumbled fstab) left it prompting for the root password without a working keyboard. Well, at least now the blasted driver's compiled in.
Re:USB drivers still in kernel? (Score:4, Informative)
Amusingly, NT4 is where they merged the Kernel and GDI memory spaces in pursuit of graphics performance. Well, they got it, but they also absolutely destroyed NT's reliablity. 3.51 was a rock. Granted, a rock with a 2GB filesystem limit...
Re: (Score:2)
Mine uses the Win2000 drivers
Re: (Score:1)
> Is there another OS with USB drivers not loaded into kernel space?
Minix?
Kernel/User: Linux is mixed (Score:3)
Actually it's both.
You can write a .ko that will be loaded by the kernel to handle your device
(used on most Linux for a few things where speed matters, like mass storage, network.
or for booting simplicity like mouse/keyboard/bluetooth)
Or you can write an user space device that communicates with the raw USB device using libusb.
(used on the huge variant zoo of non critical USB devices, like scanners, firmware upgrader, etc.)
Re: (Score:2)
Re: (Score:2)
Technically if you have your kernel offer PCI bus access to userspace you could drive the USB host controller completely from there. Not that it would necessary be a good idea, but it would reduce the attack surface to the PCI driver/bus logic (as well as introducing a new potential security problem from userspace)
Re: (Score:2)
DMA makes that approach a nonstarter unless you have a working and properly configured IOMMU between the controller and main memory. Even then, the most common use case is to give a virtual machine direct access to a device rather than to put an ordinary driver in user space.
Typo (Score:3)
Re: Typo (Score:1)
Yeah, TBL flushing causes the 4.14 to change to 4.11
It's a bug, to fixed in 4.04
Re:Typo (Score:5, Interesting)
... and PCID is not an instruction. The feature means that there is a "process ID" tag on each entry in the TLB to avoid having to flush them unnecessarily.
The intended benefit is that all entries would not necessarily have to be reloaded from page tables in RAM (or cache) whenever there is a context switch.
"Tagged TLB"s have been available on other CPU architectures for decades -- and have been used by the Linux kernels for those architectures. The feature is pretty recent on Intel x86 CPUs though.
Correct me if I'm mistaken but I think AMD's x86 CPUs do not have PCID specifically but has support for "virtual machine ID" tags on the hypervisor's second-level TLB.
Re: (Score:2)
Re: (Score:2)
They are doing what they always do. [dilbert.com]
Re: my experience with linux (Score:1)
VB!!!!!!
The story was finished the minute it you said VB programmer setting up Linux servers....,
Comedy just happens
Re: (Score:2, Offtopic)
Re: (Score:1)
Nice troll. Were you using VB6, VBA, VBScript, or VB.Net?
Visual basic for DOS.
Re: (Score:2)
Are you stuck in 1999?
Today more than 90% of the Fortune 500 rely on Linux in some aspect
http://fortune.com/2013/05/06/... [fortune.com]
Linux 79%, Windows 39%
http://www.zdnet.com/article/l... [zdnet.com]
Even Microsoft has given in, SQL Server can now run on Linux.
Re: (Score:2)
Not quite as far back as 1999, but close. This is from 2002.
https://arstechnica.com/civis/... [arstechnica.com]
Re: (Score:2)
I should have noticed "gcc 3.1"
Re: (Score:2)
Just out of curiosity... how long ago was this?
(Given that GCC 3.1 dates from 2002 or thereabouts...)
Hopefully, it won't ... (Score:2)
... come back.
Have they fixed btrfs raid 5/6 yet? (Score:1)
Or do you still have a good chance of losing all your data when a drive fails after you've replaced one?
Bigger memory limits (Score:3)
"Original x86-64 was limited by 4-level paging to 256 TiB of virtual address space and 64 TiB of physical address space. People are already bumping into this limit: some vendors offers servers with 64 TiB of memory today. "
64TB RAM... fuck.
Re: (Score:1)
Re: (Score:1)
With that much ram, one could play a really kick-ass game of Pong.
Re: (Score:2, Interesting)
Or use Firefox.
Re: (Score:2)
Re: (Score:2)
Actually, that's probably true, though there might be corner cases...
The thing is, when you've got that much addressable space you should probably be doing paging with an LRU cache flush to an intermediate level of memory, which should save itself to disk in idle moments. This would take about 64 bits/block. One bit for "changed since read from/written to disk" and a bunch for "time of last access".
OTOH, I have my doubts that they actually have 64TB of RAM. I expect they just have a memory-mapped disk wi
Re: (Score:2, Informative)
> OTOH, I have my doubts that they actually have 64TB of RAM.
I don't. It seems that you can fit 12TB of RAM (128GB*96) into this fairly standard high-end server: http://www.dell.com/en-us/work/shop/povw/poweredge-r930 . I expect that there are niche vendors that sell absolutely _massive_ machines for people who absolutely _must_ work with huge datasets in memory.
Re: (Score:2)
Perhaps I misunderstand the assertion. Perhaps you are measuring total RAM rather than directly addressable fast RAM. If that's so, then the rest of this response is inappropriate,
But my first reaction was as follows:
Well, if you amend that to "people who feel the must", I'll agree, but virtual memory means that this is a silly attitude. I don't believe that anybody is actively working with 64TB of data at once (i.e., within, say, the same second). If they think you need that then either they're wrong o
Re: (Score:1)
That's almost enough for Chrome.
Linux 4? (Score:2)
Guys, you need to pick up the pace a bit! Chrome is at already at 61.0.3163.100 !
Eh, while trying to make this joke, Chrome told me an update was ready to install and it's now at 62.0.3202.89
Re: (Score:2)
Re: (Score:2)
If found the incident of Chrome updating while I was in the middle of doing a joke related to Chrome's stupidly high version to be funnier than my original joke.
Systemd? (Score:1, Insightful)
Have they deliberately disabled all Systemd compatibility yet?
Re: (Score:2)
systemd broke kernel land, who gives a shit any more after that?
Re: (Score:1)
Not a fan of systemd either, but when userland breaks "kernelland" then the problem is with kernel not userland.
Umm... (Score:4, Informative)
128PiB of virtual address space, 4PiB of physical (Score:1)
Turns out they've just added another level to the page tables, taking it to 5.
https://www.kernel.org/doc/Doc... [kernel.org]
https://software.intel.com/sit... [intel.com]
I.e. looking up a virtual address now needs a lookup in PML5, PML4, Page Directory, Page Table. Of course the TLB caches lookups but adding more layers increases the time taken to handle a TLB miss.
I was hoping either Intel or AMD would introduce a more advanced page table - hashed inverted page tables like the ones used in PowerPC, the UltraSPARC and the I