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] [thread-next>] [day] [month] [year] [list]
Message-ID: <CAD=FV=X6oS+v-ZvJWsHYAnfVg0Lurt8z4aYaboDKkib=p4zyJw@mail.gmail.com>
Date: Fri, 8 Nov 2024 16:31:12 -0800
From: Doug Anderson <dianders@...omium.org>
To: Nir Lichtman <nir@...htman.org>
Cc: jason.wessel@...driver.com, daniel.thompson@...aro.org, 
	linux-kernel@...r.kernel.org
Subject: Re: [PATCH] kdb: fix ctrl+e/a/f/b/d/p/n broken in keyboard mode

Hi,

On Fri, Nov 8, 2024 at 2:11 PM Nir Lichtman <nir@...htman.org> wrote:
>
> Problem: When using KDB via keyboard it does not react to control
> characters which are supported in serial mode.
>
> Example: Chords such as CTRL+A/E/D/P do not work in keyboard mode
>
> Solution: Before disregarding a non-printable key character, check if it
> is one of the supported control characters, I have took the control
> characters from the switch case upwards in this function that translates
> scan codes of arrow keys/backspace/home/.. to the control characters.
> I have took them all besides the TAB control character translation (I am
> not sure what that maps to on the keyboard)

Tab translates to 9, so the 9th character. Ctrl-I.


>  kernel/debug/kdb/kdb_keyboard.c | 3 +++
>  1 file changed, 3 insertions(+)
>
> diff --git a/kernel/debug/kdb/kdb_keyboard.c b/kernel/debug/kdb/kdb_keyboard.c
> index 3c2987f46f6e..2c004abd5375 100644
> --- a/kernel/debug/kdb/kdb_keyboard.c
> +++ b/kernel/debug/kdb/kdb_keyboard.c
> @@ -172,6 +172,9 @@ int kdb_get_kbd_char(void)
>         switch (KTYP(keychar)) {
>         case KT_LETTER:
>         case KT_LATIN:
> +               if (keychar == 4 || keychar == 1 || keychar == 5 || keychar == 2 ||
> +                   keychar == 16 || keychar == 14 || keychar == 6)
> +                       return keychar;         /* non-printable supported control characters (e.g. CTRL+A) */

This is probably OK, but IMO readability here (and above) could be
improved. Untested, but I think:

#define CTRL(c) (c - 64)

Then:

/* Allowlist supported control characters */
switch(keychar) {
case CTRL('A'): /* Home */
case CTRL('B'): /* Left */
case CTRL('D'): /* Del */
case CTRL('E'): /* End */
case CTRL('F'): /* Right */
case CTRL('I'): /* Tab */
case CTRL('N'): /* Down */
case CTRL('P'): /* Up */
    return keychar;
}

The code above could also use them, AKA (untested):

/* Translate special keys to equivalent Ctrl characters */
switch (scancode) {
case 0xF: /* Tab */
  return CTRL('I');
case 0x53: /* Del */
  return CTRL('D');
case 0x47: /* Home */
  return CTRL('A');
case 0x4F: /* End */
  return CTRL('E');
case 0x4B: /* Left */
  return CTRL('B');
case 0x48: /* Up */
  return CTRL('P');
case 0x50: /* Down */
  return CTRL('N');
case 0x4D: /* Right */
  return CTRL('F');
}

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ