Date: Sun, 18 Oct 2009 21:26:00 +0300 From: Blue Swirl <blauwirbel@gmail.com> To: Juergen Lock <nox@jelal.kn-bremen.de> Cc: freebsd-emulation@freebsd.org, qemu-devel@nongnu.org, Aleksej Saushev <asau@inbox.ru>, Toni <tonygio04@gmail.com> Subject: Re: [Qemu-devel] Re: playing with qemu usermode emulation on FreeBSD... Message-ID: <f43fc5580910181126l4b8eeefbj215a13d7e6b71b59@mail.gmail.com> In-Reply-To: <20091016223426.GA54110@triton8.kn-bremen.de> References: <20091007220549.GA65997@triton8.kn-bremen.de> <20091011221840.GA55502@triton8.kn-bremen.de> <f43fc5580910121255i626e8c53t3ada32be0aa25a8f@mail.gmail.com> <20091012222058.GA43121@triton8.kn-bremen.de> <20091013221932.GA32808@triton8.kn-bremen.de> <20091016223426.GA54110@triton8.kn-bremen.de>
next in thread | previous in thread | raw e-mail | index | archive | help
On Sat, Oct 17, 2009 at 1:34 AM, Juergen Lock <nox@jelal.kn-bremen.de> wrot= e: > On Wed, Oct 14, 2009 at 12:19:32AM +0200, Juergen Lock wrote: >> On Tue, Oct 13, 2009 at 12:20:58AM +0200, Juergen Lock wrote: >> > On Mon, Oct 12, 2009 at 10:55:24PM +0300, Blue Swirl wrote: >> > > On Mon, Oct 12, 2009 at 1:18 AM, Juergen Lock <nox@jelal.kn-bremen.d= e> wrote: >> > > > On Thu, Oct 08, 2009 at 12:05:49AM +0200, Juergen Lock wrote: >> > > >> I recently noticed there are x86 bsd-user targets now (yeah I tot= ally >> > > >> missed those commits...) and now got it working a tiny little bit= : >> > > >> I can run >> > > >> =C2=A0 =C2=A0 =C2=A0 qemu-x86_64 -bsd freebsd /rescue/echo foo ba= r >> > > >> here on FreeBSD 8/amd64 and it echoes foo bar as expected, but >> > > >> segfaults afterwards. :) =C2=A0(in pthread_setcancelstate() invok= ed from >> > > >> a guest write() syscall, in case anyone is wondering.) =C2=A0Othe= r things >> > > >> I tried either exit with errors or segfault as well, and i386 hos= ts >> > > >> probably still don't work at all yet. =C2=A0(qemu-i386 here on am= d64 does >> > > >> at least something, but probably needs lock_user() treatment for = all >> > > >> kinds of syscalls, I only tried adding that for sysctl so far.) >> > > >> >> > > >> =C2=A0Anyway, here is an emulators/qemu-devel git head snapshot p= ort >> > > >> update with my current patches (files/patch-bsd-user), feel free = to >> > > >> test/debug/improve: >> > > >> =C2=A0 =C2=A0 =C2=A0 http://people.freebsd.org/~nox/qemu/qemu-dev= el-20091007.patch >> > > >> (For the folks reading this on the qemu list: =C2=A0I shall start= doing >> > > >> `proper' patch submissions later, this is more for the FreeBSD fo= lks >> > > >> and because I was asked to send what I have...) >> > > > >> > > > New version at the same place, which now runs FreeBSD/{i386,sparc6= 4} >> > > > /rescue/echo on FreeBSD/amd64, the FreeBSD/amd64 target now segfau= lts >> > > > in pthread_setcancelstate() invoked from the final writev() tho. >> > > > Oh and I also uploaded the snapshot tarball so others can now actu= ally >> > > > build the port too... :) =C2=A0And I have switched to the cpu-exec= .c patch >> > > > posted by Aleksej Saushev on the qemu list and added back amd64 >> > > > code there. >> > > > >> > > > =C2=A0Here is the bsd-user patch again: >> > > >> > > Please add Signed-off-by: line and use 'diff -u' (or preferably git = diff). >> > > >> > Well I wasn't expecting this diff to be committed just yet anyway, >> > it's still more a wip version... >> > >> > > > + =C2=A0 =C2=A0if (1 /* bsd_type =3D=3D target_freebsd */) >> > > > + =C2=A0 =C2=A0 =C2=A0 =C2=A0regs->rdi =3D infop->start_stack; >> > > >> > > Why the if and comment? >> > > >> > > > + =C2=A0 =C2=A0 =C2=A0 =C2=A0if (1 /* bsd_type =3D=3D target_freeb= sd */) { >> > > > + =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0regs->u_regs[8] =3D inf= op->start_stack; >> > > > + =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0regs->u_regs[11] =3D in= fop->start_stack; >> > > >> > > Same here. >> > > >> > =C2=A0Because bsd_type isn't available at these places in the code but >> > probably should be checked, I still wanted to fix that. =C2=A0(Maybe >> > make it global?) >> > >> I still haven't fixed this... >> >> > > > =C2=A0 =C2=A0 =C2=A0 =C2=A0 case 0x100: >> > > > + =C2=A0 =C2=A0 =C2=A0 =C2=A0/* FreeBSD uses 0x141 for syscalls to= o */ >> > > > + =C2=A0 =C2=A0 =C2=A0 =C2=A0case 0x141: >> > > > + =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0if (bsd_type !=3D targe= t_freebsd) >> > > > + =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0goto badt= rap; >> > > >> > > You are now also trapping on case 0x100 if bsd_type !=3D target_free= bsd, >> > > which probably breaks other BSDs. >> > > >> > =C2=A0Right, thats broken, the 0x141 case should come before the 0x100 >> > here of course. >> > >> =C2=A0...but this I just fixed, and I added the multiboot.S patch, and >> fixed the port's cdrom dma disable knob (files/cdrom-dma-patch). >> (And I added the cpu-exec.c whitspace fix that was already in the >> patch I posted in the BSD support thread.) >> >> =C2=A0New version at the same place, >> =C2=A0 =C2=A0 =C2=A0 http://people.freebsd.org/~nox/qemu/qemu-devel-2009= 1007.patch >> and I now also made a shar of the patched port: >> =C2=A0 =C2=A0 =C2=A0 http://people.freebsd.org/~nox/qemu/qemu-devel-2009= 1007.shar > > Updated again, among other things I added basic FreeBSD sysarch(2) > handling, fixed syscall errno return (I had added code to set the > carry bit for the x86 target before but the sign of the returned errno > was still wrong), and I finally fixed the if (1) above (made bsd_type > global.) > > =C2=A0And, I now can run FreeBSD/amd64 /bin/sh and vim on same! :) =C2=A0= (zsh > not yet tho.) > > =C2=A0Oh and Toni tested taking FreeBSD/i386's default linker script, > changing only the load address to 0x60000000 as in qemu's and, > using that as i386.ld, he now can run qemu-i386 on FreeBSD/i386 with > simple executables too... =C2=A0See files/patch-bsd-user-ld in the shar, > which I also now moved the x86_64.ld patch to that I had talked about > earlier. =C2=A0It probably can't be used everywhere as is tho since it ha= s: > =C2=A0 =C2=A0 =C2=A0 =C2=A0OUTPUT_FORMAT("elf32-i386-freebsd", "elf32-i38= 6-freebsd", > =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2=A0 =C2= =A0"elf32-i386-freebsd") > (and I also don't know if the one currently in the tree has other > features that are needed at least on Linux, any linker gurus care > to comment?) > > =C2=A0Here is the rest of the bsd-user patches again (files/patch-bsd-use= r > in the shar), if you think they are ready to commit I'm not against it > anymore :), comments are also welcome of course. Thanks, applied. I made up a short commit message.
Want to link to this message? Use this URL: <https://mail-archive.FreeBSD.org/cgi/mid.cgi?f43fc5580910181126l4b8eeefbj215a13d7e6b71b59>