linux-kernel - [stable 4.9.y 3/4] scripts/gdb: lx-dmesg: use explicit encoding=utf8 errors=replace

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

Message-Id: <20210107225229.1502459-4-f.fainelli@gmail.com>
Date:   Thu,  7 Jan 2021 14:52:28 -0800
From:   Florian Fainelli <f.fainelli@...il.com>
To:     stable@...r.kernel.org, linux-kernel@...r.kernel.org
Cc:     Leonard Crestez <leonard.crestez@....com>,
        Jan Kiszka <jan.kiszka@...mens.com>,
        Jason Wessel <jason.wessel@...driver.com>,
        Kieran Bingham <kieran@...uared.org.uk>,
        Andrew Morton <akpm@...ux-foundation.org>,
        Linus Torvalds <torvalds@...ux-foundation.org>,
        Florian Fainelli <f.fainelli@...il.com>,
        Kieran Bingham <kieran@...uared.org.uk>,
        André Draszik <git@...red.net>,
        Du Changbin <changbin.du@...il.com>
Subject: [stable 4.9.y 3/4] scripts/gdb: lx-dmesg: use explicit encoding=utf8 errors=replace

From: Leonard Crestez <leonard.crestez@....com>

commit 46d10a094353c05144f3b0530516bdac3ce7c435 upstream

Use errors=replace because it is never desirable for lx-dmesg to fail on
string decoding errors, not even if the log buffer is corrupt and we
show incorrect info.

The kernel will sometimes print utf8, for example the copyright symbol
from jffs2.  In order to make this work specify 'utf8' everywhere
because python2 otherwise defaults to 'ascii'.

In theory the second errors='replace' is not be required because
everything that can be decoded as utf8 should also be encodable back to
utf8.  But it's better to be extra safe here.  It's worth noting that
this is definitely not true for encoding='ascii', unknown characters are
replaced with U+FFFD REPLACEMENT CHARACTER and they fail to encode back
to ascii.

Link: http://lkml.kernel.org/r/acee067f3345954ed41efb77b80eebdc038619c6.1498481469.git.leonard.crestez@nxp.com
Signed-off-by: Leonard Crestez <leonard.crestez@....com>
Acked-by: Jan Kiszka <jan.kiszka@...mens.com>
Cc: Jason Wessel <jason.wessel@...driver.com>
Cc: Kieran Bingham <kieran@...uared.org.uk>
Signed-off-by: Andrew Morton <akpm@...ux-foundation.org>
Signed-off-by: Linus Torvalds <torvalds@...ux-foundation.org>
Signed-off-by: Florian Fainelli <f.fainelli@...il.com>
---
 scripts/gdb/linux/dmesg.py | 13 ++++++++++---
 1 file changed, 10 insertions(+), 3 deletions(-)

diff --git a/scripts/gdb/linux/dmesg.py b/scripts/gdb/linux/dmesg.py
index f5a030333dfd..6d2e09a2ad2f 100644
--- a/scripts/gdb/linux/dmesg.py
+++ b/scripts/gdb/linux/dmesg.py
@@ -12,6 +12,7 @@
 #
 
 import gdb
+import sys
 
 from linux import utils
 
@@ -52,13 +53,19 @@ class LxDmesg(gdb.Command):
                 continue
 
             text_len = utils.read_u16(log_buf[pos + 10:pos + 12])
-            text = log_buf[pos + 16:pos + 16 + text_len].decode()
+            text = log_buf[pos + 16:pos + 16 + text_len].decode(
+                encoding='utf8', errors='replace')
             time_stamp = utils.read_u64(log_buf[pos:pos + 8])
 
             for line in text.splitlines():
-                gdb.write("[{time:12.6f}] {line}\n".format(
+                msg = u"[{time:12.6f}] {line}\n".format(
                     time=time_stamp / 1000000000.0,
-                    line=line))
+                    line=line)
+                # With python2 gdb.write will attempt to convert unicode to
+                # ascii and might fail so pass an utf8-encoded str instead.
+                if sys.hexversion < 0x03000000:
+                    msg = msg.encode(encoding='utf8', errors='replace')
+                gdb.write(msg)
 
             pos += length
 
-- 
2.25.1