Re: [PATCH] signal: always allocate siginfo for SI_TKILL

2019-02-05 Thread Christian Brauner
On Mon, Feb 04, 2019 at 08:41:15PM -0600, Eric W. Biederman wrote: > Kees Cook writes: > > > On Sun, Feb 3, 2019 at 12:39 AM Christian Brauner > > wrote: > >> > >> On Sat, Feb 02, 2019 at 09:49:38PM -1000, Jack Andersen wrote: > >> > The patch titled > >> > `signal: Never allocate siginfo for S

Re: [PATCH] signal: always allocate siginfo for SI_TKILL

2019-02-04 Thread Eric W. Biederman
Linus Torvalds writes: > On Tue, Feb 5, 2019 at 2:41 AM Eric W. Biederman > wrote: >> >> I think the simpler change to just do: > > Ack. Changelog and sign-off? Apologies I got distracted by processes ignoring SIGKILL. I will write this up and send it out in the morning. Eric

Re: [PATCH] signal: always allocate siginfo for SI_TKILL

2019-02-04 Thread Linus Torvalds
On Tue, Feb 5, 2019 at 2:41 AM Eric W. Biederman wrote: > > I think the simpler change to just do: Ack. Changelog and sign-off? Linus

Re: [PATCH] signal: always allocate siginfo for SI_TKILL

2019-02-04 Thread Eric W. Biederman
Kees Cook writes: > On Sun, Feb 3, 2019 at 12:39 AM Christian Brauner > wrote: >> >> On Sat, Feb 02, 2019 at 09:49:38PM -1000, Jack Andersen wrote: >> > The patch titled >> > `signal: Never allocate siginfo for SIGKILL or SIGSTOP` >> > created a regression for users of PTRACE_GETSIGINFO needing

Re: [PATCH] signal: always allocate siginfo for SI_TKILL

2019-02-04 Thread Christian Brauner
On Sun, Feb 03, 2019 at 10:33:55AM -0800, Kees Cook wrote: > On Sun, Feb 3, 2019 at 12:39 AM Christian Brauner > wrote: > > > > On Sat, Feb 02, 2019 at 09:49:38PM -1000, Jack Andersen wrote: > > > The patch titled > > > `signal: Never allocate siginfo for SIGKILL or SIGSTOP` > > > created a regre

Re: [PATCH] signal: always allocate siginfo for SI_TKILL

2019-02-03 Thread Kees Cook
On Sun, Feb 3, 2019 at 12:39 AM Christian Brauner wrote: > > On Sat, Feb 02, 2019 at 09:49:38PM -1000, Jack Andersen wrote: > > The patch titled > > `signal: Never allocate siginfo for SIGKILL or SIGSTOP` > > created a regression for users of PTRACE_GETSIGINFO needing to > > discern signals that w

Re: [PATCH] signal: always allocate siginfo for SI_TKILL

2019-02-03 Thread Christian Brauner
On Sat, Feb 02, 2019 at 09:49:38PM -1000, Jack Andersen wrote: > The patch titled > `signal: Never allocate siginfo for SIGKILL or SIGSTOP` > created a regression for users of PTRACE_GETSIGINFO needing to > discern signals that were raised via the tgkill syscall. > > A notable user of this tgkill+

[PATCH] signal: always allocate siginfo for SI_TKILL

2019-02-02 Thread Jack Andersen
The patch titled `signal: Never allocate siginfo for SIGKILL or SIGSTOP` created a regression for users of PTRACE_GETSIGINFO needing to discern signals that were raised via the tgkill syscall. A notable user of this tgkill+ptrace combination is lldb while debugging a multithreaded program. Without