linux-kernel - Re: [PATCH v1] arch/x86: port I/O tracing on x86

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite for Android: free password hash cracker in your pocket

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <62596842-2389-889c-752b-632cee7a6da7@foxvalley.net>
Date:   Mon, 25 Sep 2023 15:10:20 -0600
From:   Dan Raymond <draymond@...valley.net>
To:     linux-kernel@...r.kernel.org, x86@...nel.org, mingo@...hat.com,
        bp@...en8.de, dave.hansen@...ux.intel.com, hpa@...or.com,
        Peter Zijlstra <peterz@...radead.org>
Subject: Re: [PATCH v1] arch/x86: port I/O tracing on x86

Can I get more feedback on this please?  I think I've addressed all of
Peter Zijlstra's concerns:

1) cannot use early_console->write()

   That is not true.  Everything that includes boot.h (ie.
   'arch/x86/boot/early_serial_console.c') will not change due to the
   '#ifdef BOOT_BOOT_H' guard.  Also tracing will not be compiled in
   unless 'CONFIG_TRACEPOINTS' is true.  Finally, tracing will be
   disabled at boot until the user mounts the tracefs and explicitly
   enables 'portio' events.  When tracing is disabled the trace
   routines execute a nop instruction and return immediately.

2) tracing relies on RCU which might not be alive yet and might
   itself be the problem

   RCU is not needed unless/until tracing is enabled.  I also
   demonstrated that port I/O tracing works correctly during boot by
   testing this patch with "trace_event=portio" on the command line.

I plan to submit another patch due to the compiler warning reported
by the "kernel test robot".  Before I do that I would like to get
more feedback so I can address any other concerns or suggestions.