Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

03.09.2015 01:25, Andy Lutomirski пишет: On Wed, Sep 2, 2015 at 3:25 PM, Stas Sergeev wrote: How dosemu2 is supposed to do this: 1. sigreturn() (to DOS) 2. siglongjmp() (to 64bit C-coded) This should work fine on any kernel, right? 1 - not. 2 - maybe. If, as you say, siglongjmp() restores

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Wed, Sep 2, 2015 at 3:25 PM, Stas Sergeev wrote: > 03.09.2015 00:39, Andy Lutomirski пишет: > >> On Wed, Sep 2, 2015 at 2:01 PM, Stas Sergeev wrote: >>> >>> 02.09.2015 22:06, Andy Lutomirski пишет: >>> On Wed, Sep 2, 2015 at 11:23 AM, Stas Sergeev wrote: > > 02.09.2015 21:17,

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

03.09.2015 00:39, Andy Lutomirski пишет: On Wed, Sep 2, 2015 at 2:01 PM, Stas Sergeev wrote: 02.09.2015 22:06, Andy Lutomirski пишет: On Wed, Sep 2, 2015 at 11:23 AM, Stas Sergeev wrote: 02.09.2015 21:17, Andy Lutomirski пишет: On Wed, Sep 2, 2015 at 10:46 AM, Stas Sergeev wrote:

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Wed, Sep 2, 2015 at 2:01 PM, Stas Sergeev wrote: > 02.09.2015 22:06, Andy Lutomirski пишет: > >> On Wed, Sep 2, 2015 at 11:23 AM, Stas Sergeev wrote: >>> >>> 02.09.2015 21:17, Andy Lutomirski пишет: On Wed, Sep 2, 2015 at 10:46 AM, Stas Sergeev wrote: > > 02.09.2015 17:21,

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

02.09.2015 22:06, Andy Lutomirski пишет: On Wed, Sep 2, 2015 at 11:23 AM, Stas Sergeev wrote: 02.09.2015 21:17, Andy Lutomirski пишет: On Wed, Sep 2, 2015 at 10:46 AM, Stas Sergeev wrote: 02.09.2015 17:21, Andy Lutomirski пишет: This should work for old DOSEMU. It's a bit gross, but it

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Wed, Sep 2, 2015 at 11:23 AM, Stas Sergeev wrote: > 02.09.2015 21:17, Andy Lutomirski пишет: >> On Wed, Sep 2, 2015 at 10:46 AM, Stas Sergeev wrote: >>> 02.09.2015 17:21, Andy Lutomirski пишет: >> This should work for old DOSEMU. It's a bit gross, but it has the >> nice benefit that

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

02.09.2015 21:17, Andy Lutomirski пишет: > On Wed, Sep 2, 2015 at 10:46 AM, Stas Sergeev wrote: >> 02.09.2015 17:21, Andy Lutomirski пишет: > This should work for old DOSEMU. It's a bit gross, but it has the > nice benefit that everyone (even things that aren't DOSEMU) gain the >

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Wed, Sep 2, 2015 at 10:46 AM, Stas Sergeev wrote: > 02.09.2015 17:21, Andy Lutomirski пишет: This should work for old DOSEMU. It's a bit gross, but it has the nice benefit that everyone (even things that aren't DOSEMU) gain the ability to catch signals thrown from bogus SS

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

02.09.2015 17:21, Andy Lutomirski пишет: >>> This should work for old DOSEMU. It's a bit gross, but it has the >>> nice benefit that everyone (even things that aren't DOSEMU) gain the >>> ability to catch signals thrown from bogus SS contexts, which probably >>> improves debugability. It's also

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Wed, Sep 2, 2015 at 7:21 AM, Andy Lutomirski wrote: > On Wed, Sep 2, 2015 at 2:17 AM, Stas Sergeev wrote: >> 02.09.2015 08:12, Andy Lutomirski пишет: >> >>> On Wed, Aug 19, 2015 at 9:30 AM, Stas Sergeev wrote: 19.08.2015 18:46, Andy Lutomirski пишет: > > On Wed, Aug 19,

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Wed, Sep 2, 2015 at 2:17 AM, Stas Sergeev wrote: > 02.09.2015 08:12, Andy Lutomirski пишет: > >> On Wed, Aug 19, 2015 at 9:30 AM, Stas Sergeev wrote: >>> >>> 19.08.2015 18:46, Andy Lutomirski пишет: On Wed, Aug 19, 2015 at 2:35 AM, Stas Sergeev wrote: >> >> Incidentally, I

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

02.09.2015 08:12, Andy Lutomirski пишет: On Wed, Aug 19, 2015 at 9:30 AM, Stas Sergeev wrote: 19.08.2015 18:46, Andy Lutomirski пишет: On Wed, Aug 19, 2015 at 2:35 AM, Stas Sergeev wrote: Incidentally, I tried implementing the sigaction flag approach. I think it's no good. When we return

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

02.09.2015 08:12, Andy Lutomirski пишет: On Wed, Aug 19, 2015 at 9:30 AM, Stas Sergeev wrote: 19.08.2015 18:46, Andy Lutomirski пишет: On Wed, Aug 19, 2015 at 2:35 AM, Stas Sergeev wrote: Incidentally, I tried implementing the sigaction flag approach. I think

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

02.09.2015 17:21, Andy Lutomirski пишет: >>> This should work for old DOSEMU. It's a bit gross, but it has the >>> nice benefit that everyone (even things that aren't DOSEMU) gain the >>> ability to catch signals thrown from bogus SS contexts, which probably >>> improves debugability. It's also

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

02.09.2015 21:17, Andy Lutomirski пишет: > On Wed, Sep 2, 2015 at 10:46 AM, Stas Sergeev wrote: >> 02.09.2015 17:21, Andy Lutomirski пишет: > This should work for old DOSEMU. It's a bit gross, but it has the > nice benefit that everyone (even things that aren't DOSEMU) gain

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Wed, Sep 2, 2015 at 10:46 AM, Stas Sergeev wrote: > 02.09.2015 17:21, Andy Lutomirski пишет: This should work for old DOSEMU. It's a bit gross, but it has the nice benefit that everyone (even things that aren't DOSEMU) gain the ability to catch signals thrown from

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Wed, Sep 2, 2015 at 11:23 AM, Stas Sergeev wrote: > 02.09.2015 21:17, Andy Lutomirski пишет: >> On Wed, Sep 2, 2015 at 10:46 AM, Stas Sergeev wrote: >>> 02.09.2015 17:21, Andy Lutomirski пишет: >> This should work for old DOSEMU. It's a bit gross, but it has

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

02.09.2015 22:06, Andy Lutomirski пишет: On Wed, Sep 2, 2015 at 11:23 AM, Stas Sergeev wrote: 02.09.2015 21:17, Andy Lutomirski пишет: On Wed, Sep 2, 2015 at 10:46 AM, Stas Sergeev wrote: 02.09.2015 17:21, Andy Lutomirski пишет: This should work for old DOSEMU.

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

03.09.2015 01:25, Andy Lutomirski пишет: On Wed, Sep 2, 2015 at 3:25 PM, Stas Sergeev wrote: How dosemu2 is supposed to do this: 1. sigreturn() (to DOS) 2. siglongjmp() (to 64bit C-coded) This should work fine on any kernel, right? 1 - not. 2 - maybe. If, as you say,

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Wed, Sep 2, 2015 at 2:01 PM, Stas Sergeev wrote: > 02.09.2015 22:06, Andy Lutomirski пишет: > >> On Wed, Sep 2, 2015 at 11:23 AM, Stas Sergeev wrote: >>> >>> 02.09.2015 21:17, Andy Lutomirski пишет: On Wed, Sep 2, 2015 at 10:46 AM, Stas Sergeev

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

03.09.2015 00:39, Andy Lutomirski пишет: On Wed, Sep 2, 2015 at 2:01 PM, Stas Sergeev wrote: 02.09.2015 22:06, Andy Lutomirski пишет: On Wed, Sep 2, 2015 at 11:23 AM, Stas Sergeev wrote: 02.09.2015 21:17, Andy Lutomirski пишет: On Wed, Sep 2, 2015 at 10:46 AM,

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Wed, Sep 2, 2015 at 3:25 PM, Stas Sergeev wrote: > 03.09.2015 00:39, Andy Lutomirski пишет: > >> On Wed, Sep 2, 2015 at 2:01 PM, Stas Sergeev wrote: >>> >>> 02.09.2015 22:06, Andy Lutomirski пишет: >>> On Wed, Sep 2, 2015 at 11:23 AM, Stas Sergeev

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Wed, Sep 2, 2015 at 7:21 AM, Andy Lutomirski wrote: > On Wed, Sep 2, 2015 at 2:17 AM, Stas Sergeev wrote: >> 02.09.2015 08:12, Andy Lutomirski пишет: >> >>> On Wed, Aug 19, 2015 at 9:30 AM, Stas Sergeev wrote: 19.08.2015 18:46, Andy

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Wed, Sep 2, 2015 at 2:17 AM, Stas Sergeev wrote: > 02.09.2015 08:12, Andy Lutomirski пишет: > >> On Wed, Aug 19, 2015 at 9:30 AM, Stas Sergeev wrote: >>> >>> 19.08.2015 18:46, Andy Lutomirski пишет: On Wed, Aug 19, 2015 at 2:35 AM, Stas Sergeev

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Wed, Aug 19, 2015 at 9:30 AM, Stas Sergeev wrote: > 19.08.2015 18:46, Andy Lutomirski пишет: >> On Wed, Aug 19, 2015 at 2:35 AM, Stas Sergeev wrote: Incidentally, I tried implementing the sigaction flag approach. I think it's no good. When we return from a signal, there's no

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Wed, Aug 19, 2015 at 9:30 AM, Stas Sergeev wrote: > 19.08.2015 18:46, Andy Lutomirski пишет: >> On Wed, Aug 19, 2015 at 2:35 AM, Stas Sergeev wrote: Incidentally, I tried implementing the sigaction flag approach. I think it's no good. When we return

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

* Stas Sergeev wrote: > Also, the fact that dosemu already have that functionality, > doesn't mean it will not use the new API - it actually will. So if dosemu makes use of the new facility then sure, I'm not against it at all! Thanks, Ingo -- To unsubscribe from this list: send the

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

* Stas Sergeev s...@list.ru wrote: Also, the fact that dosemu already have that functionality, doesn't mean it will not use the new API - it actually will. So if dosemu makes use of the new facility then sure, I'm not against it at all! Thanks, Ingo -- To unsubscribe from this list:

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

22.08.2015 15:38, Ingo Molnar пишет: [...] We could add yet more heuristics and teach sigreturn to ignore the saved SS value in sigcontext if the saved CS is 64-bit and the saved SS is unusable. We could maybe try this - assuming it doesn't break DOSEMU. Or we could extend the ABI and allow

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

* Andy Lutomirski wrote: > > The crash happens when DOS program terminates. > > At that point dosemu subverts the execution flow by > > replacing segregs and cs/ip ss/sp in sigcontext with its own. > > But __pad0 still has DOS SS, which crash because (presumably) > > the DOS LDT have been just

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

* Andy Lutomirski l...@amacapital.net wrote: The crash happens when DOS program terminates. At that point dosemu subverts the execution flow by replacing segregs and cs/ip ss/sp in sigcontext with its own. But __pad0 still has DOS SS, which crash because (presumably) the DOS LDT have

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

22.08.2015 15:38, Ingo Molnar пишет: [...] We could add yet more heuristics and teach sigreturn to ignore the saved SS value in sigcontext if the saved CS is 64-bit and the saved SS is unusable. We could maybe try this - assuming it doesn't break DOSEMU. Or we could extend the ABI and allow

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

19.08.2015 18:46, Andy Lutomirski пишет: > On Wed, Aug 19, 2015 at 2:35 AM, Stas Sergeev wrote: >>> Incidentally, I tried implementing the sigaction flag approach. I >>> think it's no good. When we return from a signal, there's no concept >>> of sigaction -- it's just sigreturn. Sigreturn

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Wed, Aug 19, 2015 at 2:35 AM, Stas Sergeev wrote: > 19.08.2015 01:47, Andy Lutomirski пишет: >> On Mon, Aug 17, 2015 at 11:19 PM, Stas Sergeev wrote: >>> 14.08.2015 04:37, Andy Lutomirski пишет: >>> On Thu, Aug 13, 2015 at 6:32 PM, Stas Sergeev wrote: > > 14.08.2015 04:21, Andy

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Wed, Aug 19, 2015 at 3:10 AM, Stas Sergeev wrote: > 19.08.2015 01:42, Andy Lutomirski пишет: >> What do you mean lack of proper 32/16 bit support? > At least the following: > > 1. vm86(). > There was a patch: > http://v86-64.sourceforge.net/ > Afaik rejected by Andi Kleen (likely for a good

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

19.08.2015 01:42, Andy Lutomirski пишет: > On Mon, Aug 17, 2015 at 11:29 PM, Stas Sergeev wrote: >> 13.08.2015 20:00, Brian Gerst пишет: >> >>> On Thu, Aug 13, 2015 at 11:43 AM, Andy Lutomirski >>> wrote: On Thu, Aug 13, 2015 at 8:37 AM, Linus Torvalds wrote: > > On Tue,

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

19.08.2015 01:47, Andy Lutomirski пишет: > On Mon, Aug 17, 2015 at 11:19 PM, Stas Sergeev wrote: >> 14.08.2015 04:37, Andy Lutomirski пишет: >> >>> On Thu, Aug 13, 2015 at 6:32 PM, Stas Sergeev wrote: 14.08.2015 04:21, Andy Lutomirski пишет: > On Thu, Aug 13, 2015 at 5:50 PM,

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Wed, Aug 19, 2015 at 3:10 AM, Stas Sergeev s...@list.ru wrote: 19.08.2015 01:42, Andy Lutomirski пишет: What do you mean lack of proper 32/16 bit support? At least the following: 1. vm86(). There was a patch: http://v86-64.sourceforge.net/ Afaik rejected by Andi Kleen (likely for a good

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Wed, Aug 19, 2015 at 2:35 AM, Stas Sergeev s...@list.ru wrote: 19.08.2015 01:47, Andy Lutomirski пишет: On Mon, Aug 17, 2015 at 11:19 PM, Stas Sergeev s...@list.ru wrote: 14.08.2015 04:37, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 6:32 PM, Stas Sergeev s...@list.ru wrote: 14.08.2015

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

19.08.2015 18:46, Andy Lutomirski пишет: On Wed, Aug 19, 2015 at 2:35 AM, Stas Sergeev s...@list.ru wrote: Incidentally, I tried implementing the sigaction flag approach. I think it's no good. When we return from a signal, there's no concept of sigaction -- it's just sigreturn. Sigreturn

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

19.08.2015 01:47, Andy Lutomirski пишет: On Mon, Aug 17, 2015 at 11:19 PM, Stas Sergeev s...@list.ru wrote: 14.08.2015 04:37, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 6:32 PM, Stas Sergeev s...@list.ru wrote: 14.08.2015 04:21, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 5:50 PM,

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

19.08.2015 01:42, Andy Lutomirski пишет: On Mon, Aug 17, 2015 at 11:29 PM, Stas Sergeev s...@list.ru wrote: 13.08.2015 20:00, Brian Gerst пишет: On Thu, Aug 13, 2015 at 11:43 AM, Andy Lutomirski l...@amacapital.net wrote: On Thu, Aug 13, 2015 at 8:37 AM, Linus Torvalds

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Mon, Aug 17, 2015 at 11:19 PM, Stas Sergeev wrote: > 14.08.2015 04:37, Andy Lutomirski пишет: > >> On Thu, Aug 13, 2015 at 6:32 PM, Stas Sergeev wrote: >>> >>> 14.08.2015 04:21, Andy Lutomirski пишет: >>> On Thu, Aug 13, 2015 at 5:50 PM, Stas Sergeev wrote: > > 14.08.2015 03:27,

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Mon, Aug 17, 2015 at 11:29 PM, Stas Sergeev wrote: > 13.08.2015 20:00, Brian Gerst пишет: > >> On Thu, Aug 13, 2015 at 11:43 AM, Andy Lutomirski >> wrote: >>> >>> On Thu, Aug 13, 2015 at 8:37 AM, Linus Torvalds >>> wrote: On Tue, Aug 11, 2015 at 5:17 PM, Stas Sergeev wrote: >

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

13.08.2015 23:07, Linus Torvalds пишет: On Thu, Aug 13, 2015 at 12:59 PM, Stas Sergeev wrote: It doesn't: fedora provides a "sanitized up" version of sigcontext.h in /usr/include/bits, which comes from glibc-headers-2.21-7.fc22.x86_64. So it seems the "sanitized up" headers come from glibc,

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

13.08.2015 20:00, Brian Gerst пишет: On Thu, Aug 13, 2015 at 11:43 AM, Andy Lutomirski wrote: On Thu, Aug 13, 2015 at 8:37 AM, Linus Torvalds wrote: On Tue, Aug 11, 2015 at 5:17 PM, Stas Sergeev wrote: I realize this patch may be good to have in general, but breaking userspace without a

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

14.08.2015 04:37, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 6:32 PM, Stas Sergeev wrote: 14.08.2015 04:21, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 5:50 PM, Stas Sergeev wrote: 14.08.2015 03:27, Linus Torvalds пишет: On Thu, Aug 13, 2015 at 5:17 PM, Stas Sergeev wrote: For

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

13.08.2015 20:00, Brian Gerst пишет: On Thu, Aug 13, 2015 at 11:43 AM, Andy Lutomirski l...@amacapital.net wrote: On Thu, Aug 13, 2015 at 8:37 AM, Linus Torvalds torva...@linux-foundation.org wrote: On Tue, Aug 11, 2015 at 5:17 PM, Stas Sergeev s...@list.ru wrote: I realize this patch may be

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

14.08.2015 04:37, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 6:32 PM, Stas Sergeev s...@list.ru wrote: 14.08.2015 04:21, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 5:50 PM, Stas Sergeev s...@list.ru wrote: 14.08.2015 03:27, Linus Torvalds пишет: On Thu, Aug 13, 2015 at 5:17 PM, Stas

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

13.08.2015 23:07, Linus Torvalds пишет: On Thu, Aug 13, 2015 at 12:59 PM, Stas Sergeev s...@list.ru wrote: It doesn't: fedora provides a sanitized up version of sigcontext.h in /usr/include/bits, which comes from glibc-headers-2.21-7.fc22.x86_64. So it seems the sanitized up headers come from

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Mon, Aug 17, 2015 at 11:19 PM, Stas Sergeev s...@list.ru wrote: 14.08.2015 04:37, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 6:32 PM, Stas Sergeev s...@list.ru wrote: 14.08.2015 04:21, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 5:50 PM, Stas Sergeev s...@list.ru wrote:

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Mon, Aug 17, 2015 at 11:29 PM, Stas Sergeev s...@list.ru wrote: 13.08.2015 20:00, Brian Gerst пишет: On Thu, Aug 13, 2015 at 11:43 AM, Andy Lutomirski l...@amacapital.net wrote: On Thu, Aug 13, 2015 at 8:37 AM, Linus Torvalds torva...@linux-foundation.org wrote: On Tue, Aug 11, 2015 at

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Fri, Aug 14, 2015 at 01:02:14PM +0300, Pavel Emelyanov wrote: ... > > IOW we've been not setting up __pad0 which became ss > > inside the kernel (in result we've been passing 0 here, > > which caused the problem). > > > > fwiw, we declare that new criu versions may require new > > kernels to

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On 08/14/2015 10:22 AM, Cyrill Gorcunov wrote: > On Thu, Aug 13, 2015 at 01:09:47PM -0700, Linus Torvalds wrote: >> On Thu, Aug 13, 2015 at 1:08 PM, Cyrill Gorcunov wrote: >>> >>> If only I'm not missin something obvious this should not hurt us. >>> But I gonna build test kernel and check to be

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 08:43:24AM -0700, Andy Lutomirski wrote: ... > > > > That rule hasn't gone anywhere. > > > > Does a plain revert just fix everything? Because if so, that's the > > right thing to do, and we can just re-visit this later. > > > > I don't understand why Andy and Ingo are even

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 01:09:47PM -0700, Linus Torvalds wrote: > On Thu, Aug 13, 2015 at 1:08 PM, Cyrill Gorcunov wrote: > > > > If only I'm not missin something obvious this should not hurt us. > > But I gonna build test kernel and check to be sure tomorrow, ok? Managed to test it. And criu

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 01:09:47PM -0700, Linus Torvalds wrote: On Thu, Aug 13, 2015 at 1:08 PM, Cyrill Gorcunov gorcu...@gmail.com wrote: If only I'm not missin something obvious this should not hurt us. But I gonna build test kernel and check to be sure tomorrow, ok? Managed to test it.

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 08:43:24AM -0700, Andy Lutomirski wrote: ... That rule hasn't gone anywhere. Does a plain revert just fix everything? Because if so, that's the right thing to do, and we can just re-visit this later. I don't understand why Andy and Ingo are even discussing

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Fri, Aug 14, 2015 at 01:02:14PM +0300, Pavel Emelyanov wrote: ... IOW we've been not setting up __pad0 which became ss inside the kernel (in result we've been passing 0 here, which caused the problem). fwiw, we declare that new criu versions may require new kernels to work but

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On 08/14/2015 10:22 AM, Cyrill Gorcunov wrote: On Thu, Aug 13, 2015 at 01:09:47PM -0700, Linus Torvalds wrote: On Thu, Aug 13, 2015 at 1:08 PM, Cyrill Gorcunov gorcu...@gmail.com wrote: If only I'm not missin something obvious this should not hurt us. But I gonna build test kernel and check

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

14.08.2015 04:37, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 6:32 PM, Stas Sergeev wrote: 14.08.2015 04:21, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 5:50 PM, Stas Sergeev wrote: 14.08.2015 03:27, Linus Torvalds пишет: On Thu, Aug 13, 2015 at 5:17 PM, Stas Sergeev wrote: For

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 6:32 PM, Stas Sergeev wrote: > 14.08.2015 04:21, Andy Lutomirski пишет: > >> On Thu, Aug 13, 2015 at 5:50 PM, Stas Sergeev wrote: >>> >>> 14.08.2015 03:27, Linus Torvalds пишет: On Thu, Aug 13, 2015 at 5:17 PM, Stas Sergeev wrote: > > For example

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

14.08.2015 04:21, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 5:50 PM, Stas Sergeev wrote: 14.08.2015 03:27, Linus Torvalds пишет: On Thu, Aug 13, 2015 at 5:17 PM, Stas Sergeev wrote: For example because you can as well do: prctl(ARCH_SET_SIGNAL_SS, 0) which will mean "restore ss in

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 5:50 PM, Stas Sergeev wrote: > 14.08.2015 03:27, Linus Torvalds пишет: >> >> On Thu, Aug 13, 2015 at 5:17 PM, Stas Sergeev wrote: >>> >>> For example because you can as well do: >>> prctl(ARCH_SET_SIGNAL_SS, 0) >>> which will mean "restore ss in sighandler to its current

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

14.08.2015 03:27, Linus Torvalds пишет: On Thu, Aug 13, 2015 at 5:17 PM, Stas Sergeev wrote: For example because you can as well do: prctl(ARCH_SET_SIGNAL_SS, 0) which will mean "restore ss in sighandler to its current value", I really think a prctl() is the wrong thing to do. If you want a

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 5:24 PM, Andy Lutomirski wrote: > > So yes, it mostly works. It also sucks, and it makes it extremely > unpleasant for any other program to do this. Well, I'd argue that (a) we don't really _want_ any other programs to do that (b) but yeah, we might want to make it

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 5:17 PM, Stas Sergeev wrote: > > For example because you can as well do: > prctl(ARCH_SET_SIGNAL_SS, 0) > which will mean "restore ss in sighandler to its current value", I really think a prctl() is the wrong thing to do. If you want a signal handler to save/restore

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 5:08 PM, Linus Torvalds wrote: > On Thu, Aug 13, 2015 at 5:00 PM, Stas Sergeev wrote: >> 14.08.2015 02:00, Andy Lutomirski пишет: >>> >>> DOSEMU, when you set that flag, WRFSBASE gets enabled, and glibc's >>> threading library starts using WRFSBASE instead of arch_prctl.

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

14.08.2015 03:05, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 5:00 PM, Stas Sergeev wrote: 14.08.2015 02:00, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 3:51 PM, Stas Sergeev wrote: 14.08.2015 01:29, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 3:25 PM, Stas Sergeev wrote:

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 5:00 PM, Stas Sergeev wrote: > 14.08.2015 02:00, Andy Lutomirski пишет: >> >> DOSEMU, when you set that flag, WRFSBASE gets enabled, and glibc's >> threading library starts using WRFSBASE instead of arch_prctl. > > Hmm, how about the following: > >

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 5:00 PM, Stas Sergeev wrote: > 14.08.2015 02:00, Andy Lutomirski пишет: > >> On Thu, Aug 13, 2015 at 3:51 PM, Stas Sergeev wrote: >>> >>> 14.08.2015 01:29, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 3:25 PM, Stas Sergeev wrote: > > 14.08.2015 01:11,

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 4:43 PM, Stas Sergeev wrote: > In fact, in the cases I can remember, the kernel patches > were never reverted, see this for instance: > https://lkml.org/lkml/2005/3/26/21 > And there were many other breakages too, for example when > kernel started to use top-down memory

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

14.08.2015 02:00, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 3:51 PM, Stas Sergeev wrote: 14.08.2015 01:29, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 3:25 PM, Stas Sergeev wrote: 14.08.2015 01:11, Andy Lutomirski пишет: Now suppose you set some magic flag and jump (via sigreturn,

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

14.08.2015 02:18, Linus Torvalds пишет: On Thu, Aug 13, 2015 at 4:05 PM, Linus Torvalds wrote: The _only_ thing that matters is that something broke. To clarify: things like test programs etc don't matter. Real applications, used by real users. That's what regressions cover. If you have a

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On 08/13/15 16:18, Linus Torvalds wrote: On Thu, Aug 13, 2015 at 4:05 PM, Linus Torvalds wrote: The _only_ thing that matters is that something broke. To clarify: things like test programs etc don't matter. Real applications, used by real users. That's what regressions cover. If you have a

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 4:05 PM, Linus Torvalds wrote: > > The _only_ thing that matters is that something broke. To clarify: things like test programs etc don't matter. Real applications, used by real users. That's what regressions cover. If you have a workflow that isn't just some random

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

14.08.2015 02:00, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 3:51 PM, Stas Sergeev wrote: 14.08.2015 01:29, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 3:25 PM, Stas Sergeev wrote: 14.08.2015 01:11, Andy Lutomirski пишет: Now suppose you set some magic flag and jump (via sigreturn,

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 3:01 PM, Raymond Jennings wrote: > > So it still counts as a regression if the kernel pulls the rug out from > under someone that was relying on undocumented or buggy behavior? Absolutely. There are no excuses for regressions. If the code was badly written and left itself

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 3:51 PM, Stas Sergeev wrote: > 14.08.2015 01:29, Andy Lutomirski пишет: >> >> On Thu, Aug 13, 2015 at 3:25 PM, Stas Sergeev wrote: >>> >>> 14.08.2015 01:11, Andy Lutomirski пишет: >>> Now suppose you set some magic flag and jump (via sigreturn, trampoline,

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

14.08.2015 01:29, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 3:25 PM, Stas Sergeev wrote: 14.08.2015 01:11, Andy Lutomirski пишет: Now suppose you set some magic flag and jump (via sigreturn, trampoline, whatever) into DOS code. The DOS code loads 0x7 into FS and then gets #GP. You

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 3:25 PM, Stas Sergeev wrote: > 14.08.2015 01:11, Andy Lutomirski пишет: > >> Now suppose you set some magic flag and jump (via sigreturn, >> trampoline, whatever) into DOS code. The DOS code loads 0x7 into FS >> and then gets #GP. You land in a signal handler. As far as

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

14.08.2015 01:11, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 3:02 PM, Stas Sergeev wrote: 14.08.2015 00:46, Linus Torvalds пишет: On Thu, Aug 13, 2015 at 2:42 PM, Raymond Jennings wrote: I am curious about what's supposed to happen normally on signal delivery. Is SS a register that's

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 3:02 PM, Stas Sergeev wrote: > 14.08.2015 00:46, Linus Torvalds пишет: >> >> On Thu, Aug 13, 2015 at 2:42 PM, Raymond Jennings >> wrote: >>> >>> I am curious about what's supposed to happen normally on signal delivery. >>> >>> Is SS a register that's supposed to be

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

14.08.2015 01:01, Raymond Jennings пишет: On 08/13/15 14:46, Linus Torvalds wrote: On Thu, Aug 13, 2015 at 2:42 PM, Raymond Jennings wrote: I am curious about what's supposed to happen normally on signal delivery. Is SS a register that's supposed to be preserved like EIP/RIP and CS when

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

14.08.2015 00:46, Linus Torvalds пишет: On Thu, Aug 13, 2015 at 2:42 PM, Raymond Jennings wrote: I am curious about what's supposed to happen normally on signal delivery. Is SS a register that's supposed to be preserved like EIP/RIP and CS when a signal is delivered? What exactly does

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On 08/13/15 14:46, Linus Torvalds wrote: On Thu, Aug 13, 2015 at 2:42 PM, Raymond Jennings wrote: I am curious about what's supposed to happen normally on signal delivery. Is SS a register that's supposed to be preserved like EIP/RIP and CS when a signal is delivered? What exactly does

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 2:42 PM, Raymond Jennings wrote: > > I am curious about what's supposed to happen normally on signal delivery. > > Is SS a register that's supposed to be preserved like EIP/RIP and CS when a > signal is delivered? What exactly does "supposed" mean? On x86-64, we

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On 08/13/15 13:09, Linus Torvalds wrote: On Thu, Aug 13, 2015 at 1:08 PM, Cyrill Gorcunov wrote: If only I'm not missin something obvious this should not hurt us. But I gonna build test kernel and check to be sure tomorrow, ok? Thanks, Linus -- To unsubscribe from this

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

13.08.2015 22:49, Andy Lutomirski пишет: On Aug 13, 2015 12:05 PM, "Stas Sergeev" wrote: 13.08.2015 21:41, Andy Lutomirski пишет: Stas: I think uc_flags is okay. We don't currently read it during sigreturn, but I see no reason that we can't start reading it. Andy, we definitely have some

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 1:08 PM, Cyrill Gorcunov wrote: > > If only I'm not missin something obvious this should not hurt us. > But I gonna build test kernel and check to be sure tomorrow, ok? Thanks, Linus -- To unsubscribe from this list: send the line "unsubscribe

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 12:53:03PM -0700, Linus Torvalds wrote: > On Thu, Aug 13, 2015 at 11:41 AM, Andy Lutomirski wrote: > > On Thu, Aug 13, 2015 at 11:35 AM, Linus Torvalds > > wrote: > >> > >> Ok. So I'm inclined to do the bigger revert, just to fix the compile > >> issue. It would be crazy

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 12:59 PM, Stas Sergeev wrote: > > It doesn't: fedora provides a "sanitized up" version of sigcontext.h > in /usr/include/bits, which comes from glibc-headers-2.21-7.fc22.x86_64. > So it seems the "sanitized up" headers come from glibc, which > means all other distros would

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

13.08.2015 22:37, Linus Torvalds пишет: On Thu, Aug 13, 2015 at 12:13 PM, Stas Sergeev wrote: As for the compilation failure - I am surprised you even care. I thought the "we don't break userspace" covers only run-time, not compile-time. Oh well. I definitely care. Compile issues may be

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 11:41 AM, Andy Lutomirski wrote: > On Thu, Aug 13, 2015 at 11:35 AM, Linus Torvalds > wrote: >> >> Ok. So I'm inclined to do the bigger revert, just to fix the compile >> issue. It would be crazy to force some silly autoconf script for >> random header info. > > Yeah,

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Aug 13, 2015 12:05 PM, "Stas Sergeev" wrote: > > 13.08.2015 21:41, Andy Lutomirski пишет: > >> Stas: I think uc_flags is okay. We don't currently read it during >> sigreturn, but I see no reason that we can't start reading it. > > Andy, we definitely have some communication discontinuity

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 12:13 PM, Stas Sergeev wrote: > > As for the compilation failure - I am surprised you even care. > I thought the "we don't break userspace" covers only run-time, > not compile-time. Oh well. I definitely care. Compile issues may be slightly lower on my radar, but the

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

13.08.2015 22:01, Andy Lutomirski пишет: On Thu, Aug 13, 2015 at 11:57 AM, Stas Sergeev wrote: 13.08.2015 21:35, Linus Torvalds пишет: On Thu, Aug 13, 2015 at 10:51 AM, Stas Sergeev wrote: Hello Linus, I verified that patch-minimal.diff is enough to fix the problem, BUT! dosemu is in fact

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

13.08.2015 21:41, Andy Lutomirski пишет: Stas: I think uc_flags is okay. We don't currently read it during sigreturn, but I see no reason that we can't start reading it. Andy, we definitely have some communication discontinuity here. :) The point is not sigreturn. If we are talking about the

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

On Thu, Aug 13, 2015 at 11:57 AM, Stas Sergeev wrote: > 13.08.2015 21:35, Linus Torvalds пишет: >> >> On Thu, Aug 13, 2015 at 10:51 AM, Stas Sergeev wrote: >>> >>> Hello Linus, I verified that patch-minimal.diff is enough >>> to fix the problem, BUT! dosemu is in fact using the .fs and >>> .gs

Re: [regression] x86/signal/64: Fix SS handling for signals delivered to 64-bit programs breaks dosemu

13.08.2015 21:35, Linus Torvalds пишет: On Thu, Aug 13, 2015 at 10:51 AM, Stas Sergeev wrote: Hello Linus, I verified that patch-minimal.diff is enough to fix the problem, BUT! dosemu is in fact using the .fs and .gs fields of sigcontext as a placeholders. Why the minimal patch alone helps is

  1   2   3   >