I feel boring that pmap_enter(9) can not avoid to
have goto jumps for the logic simplity. This indicates
pmap_enter(9) is mistakenly designed and used for
mulitple purposes in parallel. Rework is seriously
requested..
Toru Nishimura / ALKYL Technology / Sanctum Networks
. It invovles complicated
case analysis., but in essense the case analysis is
common across processors.
Toru Nishimura / ALKYL Technology / Sanctum Networks
rather trandinal many-core processor;
64K+64K L1 caches
22MB L2 cache
16 core in 1 processor
2 HW threads per core
The inter-connect technology for multiple chasis configuration is a
Fujitsu proprietary.
Toru Nishimura / ALKYL Technology
Does anyone out there start considering to port SMP NetBSD
for the SPARC64 X products from Fujitsu ?
http://pr.fujitsu.com/jp/news/2013/01/18.html
(English PR will follow soon, I expect)
It's the least sister of Kei Super computer technology.
Toru Nishimura / ALKYL Technology
support there.
Toru Nishimura / ALKYL Technology
There are increasing number of NAND only (NOR less) embeded devices on
market. How difficult to have chewieFS LIBSA support to allow kernel image
loading from the filesys on NAND?
Toru Nishimura / ALKYL Technology
possiblity
is to have DDB symbol table as a module. This would wape out
the awesome burdern of SYMTAB_SPACE maintaintance when
LIBSA bootloader is made smart enough to do that.
Toru Nishimura / ALKYL Technology
or not)
The code segment does in-kernel socket creation and send/recv dialogue.
Toru Nishimura / ALKYL Technology
Izumi Tsutui points;
It checks a file system type where the kernel is (to be) loaded
It's a norm case that a target kernel is stored in rootfs, I think.
Toru Nishimura / ALKYL Technology
This eliminates the necessity to build a separate
INSTALL kernel which holds mdimage inside.
That's what i386 did and miniroot.kmod was loaded via boot.cfg(5)
like distrib/i386/cdroms/installcd/boot.cfg.in in netbsd-5.
(-current doesn't use miniroot.kmod and use cd9660 root fs)
It's my
since it's somehow grub'ing way to
pursue imaginary (and not-existing) universal loaders and would
limit creative imagination to florish. Scripting and even command
line editing for debug would be possible and all heavily depend
on targets to port and how to use it in field.
Toru Nishimura / ALKYL
Folks,
It would be nice if ELF kernel image holds modules_enabled value in
note section which allows LIBSA to tell whether the target kernel is
MODULAR or not.
Toru Nishimura / ALKYL Technology
matter since live PTEs are likely in cache.
TLBrefill operation is very cheap in that case as refill-target PTEs are
already fetched and stored somewhere in primary Dcache or at worst in
multi-MB L2 cache given modestly loaded system runtime.
Toru Nishimura / ALKYL Technology
different sets of bits at
* same time but these sets never overlap.
* ...
*/
Thanks for your intriguing feedfeek, matt. Do you use tlbivax insn new to
e500 for any purpose? My impression is it might not be so useful
contradicting to the intent.
Toru Nishimura / ALKYL Technology
manipuating primitives
since cache design varies radically among CPU implementations and
it would make little sense to define universal cache primitives good
enough across varying cache designs. ASID management falls into
the same category.and pmap(9) should stay away.
Toru Nishimura / ALKYL Technology
to invalidate instruction cache to continue.
Toru Nishimura / ALKYL Technology
tradeoff, I believe.
BTW, how do you approach to implement a remote TLB
shootdown?
Toru Nishimura / ALKYL Technology
Eric Haszlakiewicz said;
I doubt that would do anything useful.
...[SNIP]...
At that point you're not using the FAT FS as a filesystem, but just
as a complicated disklabel.
No better than FAT, yes, right. PLS look at plus side.
Toru Nishimura / ALKYL Technology
here
someone steps forward to implement this.
Comments and challening sprites are welcome.
Toru Nishimura / ALKYL Technology
19 matches
Mail list logo