pm215

72 ms·

811.
by pm215 10y ago
The Apache 2.0 license isn't compatible with the GPLv2 (according to the FSF's compatibility list), whereas MIT is -- that may or may not be important to you. Apache 2.0 is compatible with GPLv3, and thus also with GPLv2-or-later
812.
by pm215 10y ago
For embedded support, the platform (ie the entire SoC and its devices) is more important than just the CPU architecture. Every SoC is different and if the OS doesn't have drivers and support for the devices on it then it won't boo
813.
by pm215 10y ago
The ARM term is ASID (address space identifier). There's also a VMID so different VMs under a hypervisor can share the TLB.
814.
by pm215 10y ago
The CPU does know the physical address, because the MMU is part of the CPU; so you can have a physically indexed cache if you want. The problem is just that you can't do the cache lookup in parallel with the MMU physical-to-virtual loo
815.
by pm215 10y ago
Well, I guess technically you could do it -- the codegen started out as code from tcc, after all. However we've made enough changes over the years that right now it has some specializations to QEMU's needs. Also questions like &qu
816.
by pm215 10y ago
It's the bit where it says 'is it used in conjunction with a hypervisor?'. That's how we define the use cases that count as defendable against malicious guests. This covers more than just the TCG cpu emulation because it
817.
by pm215 10y ago
It is documented here, but you're right that ideally we could mention it somewhere more prominent. http://wiki.qemu-project.org/SecurityProcess#How_impact_and_...
818.
by pm215 10y ago
The main reason I request commits to be split up is for ease of code review. It's much easier to review three commits that each do one easily comprehensible small thing than one commit that does three things at once. It's also bet
819.
by pm215 10y ago
Asking about project xy is different though, it's more guided. I think it's much easier to answer than the really open ended ones that ask you to pick an incident from all 10-15 years of your work history: the field there is so wi
820.
by pm215 10y ago
Yeah, you need to configure it to tell it what configuration of cache you want to simulate (there's an example of doing this for the Jaguar cache in slides 84/85). Valgrind's cachegrind tool also lets you tweak the cache conf
821.
by pm215 10y ago
I think it doesn't seem a totally odd choice in the UK partly because the public transport is often pretty poor (London is a huge exception to the usual UK situation). Certainly here in Cambridge I'd choose a 6 mile cycle over the
822.
by pm215 10y ago
No, it's just that it's the kind of feature that's quite hard to retrofit to an existing design, since it breaks existing assumptions in a lot of places. And it's only relatively recently that the kind of embedded board
823.
by pm215 10y ago
I'd like to be able to use Rust in QEMU one day, but just at the moment it doesn't support all the platforms QEMU does. The sheer range supported by a standard GCC C toolchain is very hard for a new language to catch up to. (It mo
824.
by pm215 10y ago
Not only the BASIC programs you wrote (which do indeed live in VDP RAM). The BASIC interpreter itself was stored in GROM, which was read-only but had a similar slow 8-bit interface with an address port and a data port on it, and written in
825.
by pm215 10y ago
The slow BASIC was due to the completely insane doubly-interpreted implementation. The BASIC interpreter was written in GPL (Graphics Programming Language; https://en.wikipedia.org/wiki/Texas_Instruments_TI-99/4A#V
826.
by pm215 10y ago
KVM for ARM doesn't parse instructions -- you can just use the register info the hardware gives you in the syndrome register, which covers everything except oddball cases like trying load-multiple to a device, which doesn't happen
827.
by pm215 10y ago
Wait, x86 still requires instruction emulation for non-weirdo non-legacy cases? My vague recollection of the KVM Forum talk G. did was that you don't need it for "modern" guests. (We were talking about emulation-via-just-inte
828.
by pm215 10y ago
Mmm, in this case I suspect a lot of the benefit is not having to carry around the 90% of QEMU that's completely unused in the x86-KVM usecase (ie all of the emulation, all the devices for non-x86, all the random PCI devices you don&#x
829.
by pm215 10y ago
As abecedarius says, there's a good commentary on it in the Lions book ( https://en.wikipedia.org/wiki/Lions'_Commentary_on_UNIX_6th_... ). That was reprinted in the late 90s when I was at university and I read
830.
by pm215 10y ago
The cost of fork+exec of a separate binary will make even the most efficient possible external echo slower than the shell builtin, I suspect. (This is why echo is a builtin in the first place, though there's no requirement for it to be
831.
by pm215 10y ago
You reply to the email saying 'sorry, but we can't accept your code because...'. (The shortest way to do this is to just reply saying 'NAK', but I agree with Rusty Russell that nak-mails are generally unfriendly and
832.
by pm215 10y ago
32-bit ARM is a bit awkward because traditionally there is no equivalent of the x86 BIOS APIs for doing things like printing output, and the code started by a bootloader is expected to talk directly to the hardware (serial port etc), which
833.
by pm215 10y ago
ptrace() isn't undocumented, it has a fairly long manpage: http://man7.org/linux/man-pages/man2/ptrace.2.html It's certainly pretty hairy, but on the other hand only a handful of programs (debuggers
834.
by pm215 10y ago
I agree that dumping an image is not a great idea in the modern world, but the concept(1) dates right back to when emacs was still based on TECO on the PDP-10: http://pdp-10.trailing-edge.com/mit_emacs_170_teco_1220/01&
835.
by pm215 10y ago
If you're an open source project that's expanded to more than a few people, it's worth going through and making a list of all the various "admin" items your project has accumulated (website? domain name? github? bug
836.
by pm215 10y ago
You definitely can't skip the step of getting managerial and legal signoff (and no sensible open source project wants a patch that the developer doesn't have the right and authorisation from the employer to contribute). But in bet
837.
by pm215 10y ago
QEMU won't be fastest if your host is osx, because we don't support the osx virtualization API -- KVM requires a Linux host.
838.
by pm215 10y ago
Yep, I get paid to work on QEMU. I would suggest that your chances of getting paid to work on something and what kind of work that turns out to be depend quite a bit on the project. You can have a look at the git commit history or the maili
839.
by pm215 10y ago
When I first set up fail2ban it was because I got annoyed that the machine on my desk was making regular "clunk...clunk...clunk" noises from the hard disk as it wrote another failed-auth attempt to the log every second or so...
840.
by pm215 10y ago
"In 10 years" is a bit misleading, because that's just the construction time. You need to also take into account the time it took to go from "this would be a good idea" to actually starting to dig, because that&#x27

More ›