Re: [PATCH v5 0/6] Remote fork events

[Pedro Alves <palves at redhat dot com>]

Thanks Don,

On a quick skim, this version is looking much better.

I'll apply this to my sandbox and take a deeper look.

Thanks,
Pedro Alves

On 02/27/2015 12:46 AM, Don Breazeal wrote:
> This is an update to the patch series implementing remote follow-fork.
> This update only implements fork and vfork events for extended remote,
> omitting exec events and 'target remote' for now.
> 
> Changes from the previous version include:
>  * Removed 'target remote' support, which is deferred for later.
>  * Fixes per review comments, notably
>    - issues with unpushing the target after detach or kill
>    - reorganizing/renaming some things, both to match the
>      native implementation better and for just plain better
>      organization.
>    - fixed and/or improved some of the event handling code.
>  * Also added documentation to each individual patch, rather than
>    saving it all for a subsequent 'documentation patch'.
> 
> The new series includes six patches as follows:
> 
> 1/6: Preparatory patch that implements qSupported support for fork events
>      and associated mechanisms.
> 
> 2/6: Implements functions to clone breakpoint lists in gdbserver.
> 
> 3/6: Implements follow fork for 'fork' but not 'vfork', for
>      extended-remote targets only.
> 
> 4/6: Adds the architecture-specific pieces of follow-fork that allows
>      hardware watchpoints to be inherited by a forked child.
>      This patch was previously approved, although I have added a
>      trivial change to gdb/NEWS in this patch.
> 
> 5/6: Adds follow fork for 'vfork'.
> 
> 6/6: Adds catchpoints for 'fork' and 'vfork', along with support for
>      killing a process that has forked but has not been followed.
> 
> TESTING:
> Testing was mostly done using x86_64 Ubuntu, with the exception of the
> architecture-specific patch, #4.  There are a few anomalies that show
> up, but don't signify any problem.
> 
>  - Intermediate patches show failures due to the lack of features
>    implemented in subsequent patches, like missing hardware watchpoint
>    or catchpoint support.
> 
>  - Some vfork tests fail due to the lack of exec event support.
> 
>  - There are a couple of tests that show new failures that actually
>    fail in the current mainline.  Details of these are as follows:
> 
>    * when vfork events are enabled, gdb.base/disp-step-syscall.exp
>      shows PASS => FAIL in .sum diffs.  The test actually always
>      fails.  With native/master, we see
> 
>       stepi^M
>       FAIL: gdb.base/disp-step-syscall.exp: vfork: stepi vfork insn
> (timeout)
> 
>      With remote and extended-remote/master, we see a bogus PASS result:
>       stepi^M
>       [Inferior 1 (process 9399) exited normally]^M
>       (gdb) PASS: gdb.base/disp-step-syscall.exp: vfork: stepi vfork insn
> 
>     The criteria to pass that test are pretty lax:
>       gdb_test "stepi" ".*" "stepi $syscall insn"
> 
>   * Similarly, once vfork events are enabled, gdb.base/watch-vfork.exp
>     shows PASS => FAIL in .sum diffs.  This test also always fails.  With
>     native/master, we see this:
> 
>      continue^M
>      Continuing.^M
>      FAIL: gdb.base/watch-vfork.exp: Watchpoint triggers after vfork
> (sw) (timeout)
> 
>     With extended-remote/master, we see this:
>      continue
>      Continuing.
>      [Inferior 1 (process 18866) exited normally]
>      (gdb) FAIL: gdb.base/watch-vfork.exp: Watchpoint triggers after
> vfork (sw) (the program exited)
> 
>     But once vfork events are enabled, we see:
>      continue
>      Continuing.
>      Detaching from process 17405
>      FAIL: gdb.base/watch-vfork.exp: Watchpoint triggers after vfork
> (sw) (timeout)
>     which more-or-less matches the native/master behavior.
> 
> Thanks,
> --Don
>