linux-kernel - [PATCH 0/9] Macintosh II ADB driver fixes

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: Windows password security audit tool. GUI, reports in PDF.

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

Date:   Sun, 28 Jun 2020 14:23:12 +1000
From:   Finn Thain <fthain@...egraphics.com.au>
To:     Benjamin Herrenschmidt <benh@...nel.crashing.org>
Cc:     Joshua Thompson <funaho@...ai.org>,
        Geert Uytterhoeven <geert@...ux-m68k.org>,
        linux-m68k@...ts.linux-m68k.org,
        Laurent Vivier <lvivier@...hat.com>,
        Mark Cave-Ayland <mark.cave-ayland@...nde.co.uk>,
        linuxppc-dev@...ts.ozlabs.org, linux-kernel@...r.kernel.org
Subject: [PATCH 0/9] Macintosh II ADB driver fixes

Various issues with the via-macii driver have become apparent over the
years. Some examples:

 - A Talk command response can be lost. This can result in phantom devices
being probed or an incorrect device handler ID being retrieved.

 - A reply packet containing a null byte can get truncated. Such packets
are sometimes generated by ADB keyboards.

 - A Talk Register 3 reply from device 15 (that is, command byte 0xFF)
can be mistaken for a bus timeout (empty packet).

This patch series contains fixes for all known bugs in the via-macii
driver, plus a few code style improvements. It has been successfully
tested on an Apple Centris 650 and qemu-system-m68k.

The patched kernel does regress on past QEMU releases, due to ADB
transceiver emulation bugs. Those bugs have been fixed in mainline QEMU.
My thanks go to Mark Cave-Ayland for that effort and for figuring out
the improvements to the signalling between the VIA and the transceiver.

Note to -stable maintainers: these fixes can be cherry-picked without
difficulty, if you have the 5 commits that appeared in v5.0:

b52dce8738938 macintosh/via-macii: Synchronous bus reset
5f93d7081a47e macintosh/via-macii: Remove BUG_ON assertions
5ce6185c2ef4e macintosh/via-macii: Simplify locking
351e5ad327d07 macintosh/via-macii, macintosh/adb-iop: Modernize printk calls
47fd2060660e6 macintosh/via-macii, macintosh/adb-iop: Clean up whitespace

Just for the sake of simplicity, the 'fixes' tags in this series limit
backporting to 'v5.0+'.


Finn Thain (9):
  macintosh/via-macii: Access autopoll_devs when inside lock
  macintosh/via-macii: Poll the device most likely to respond
  macintosh/via-macii: Handle /CTLR_IRQ signal correctly
  macintosh/via-macii: Remove read_done state
  macintosh/via-macii: Handle poll replies correctly
  macintosh/via-macii: Use bool type for reading_reply variable
  macintosh/via-macii: Use unsigned type for autopoll_devs variable
  macintosh/via-macii: Use the stack for reset request storage
  macintosh/via-macii: Clarify definition of macii_init()

 drivers/macintosh/via-macii.c | 324 +++++++++++++++++++---------------
 1 file changed, 179 insertions(+), 145 deletions(-)

-- 
2.26.2