Author: nanxiao

The pitfall of using tshark to analyse QUIC protocol

Check the wireshark’s QUIC related code, we will find it heavily depends on 2 macros:

#ifdef HAVE_LIBGCRYPT_AEAD
......
#endif

#ifdef HAVE_LIBGCRYPT_CHACHA20
......
#endif

And these 2 macros rely on the version of libgcrypt (refer here):

/*
 * Define HAVE_LIBGCRYPT_AEAD here, because it's used in several source
 * files.
 */
#if GCRYPT_VERSION_NUMBER >= 0x010600 /* 1.6.0 */
/* Whether to provide support for authentication in addition to decryption. */
#define HAVE_LIBGCRYPT_AEAD
#endif

/*
 * Define some other "do we have?" items as well.
 */
#if GCRYPT_VERSION_NUMBER >= 0x010700 /* 1.7.0 */
/* Whether ChaCh20 PNE can be supported. */
#define HAVE_LIBGCRYPT_CHACHA20
/* Whether AEAD_CHACHA20_POLY1305 can be supported. */
#define HAVE_LIBGCRYPT_CHACHA20_POLY1305
#endif

On CentOS 7, the libgcrypt version is 1.5.3, so the above 2 macros will not be defined, and some functions are not available. While on CentOS 8, the libgcrypt version is 1.8.5, so the functions are fully supported. I met an issue, i.e., for the same pcap file, tshark (I built myself) on CentOS 7 assumes there is an error in decrypting QUIC flow:

$ /home/nanxiao/wireshark/build/run/tshark -nr 435.pcap -Y '(quic.decryption_failed)'
    1   0.000000 172.27.232.168 → 216.183.220.159 GTP <QUIC> 1310 Initial, DCID=68a3ee8706f87817

while tshark on CentOS 8 works OK:

$ /home/nanxiao/wireshark/build/run/tshark -nr 435.pcap -Y '(quic.decryption_failed)'
$

The workaround of fixing bridged networking issue with Monterey

After upgrading my macOS to Monterey, my Virtualbox’s Void Linux VM can’t connect to Internet through bridged networking, though I can still connect it via SSH terminal emulator. One workaround is changing VM’s network mode to NAT, then configure “Port Forwarding” (Refer here):
(1) Goto Settings -> Network -> Advanced -> Port Forwarding:


(2) Set forwarding port:


Then “ssh user@127.0.0.1 -p2022” should work.
BTW, for my Void Linux VM, I must modify the nameserver field in /etc/resolve.conf to a public DNS server (e.g., 8.8.8.8):

nameserver 8.8.8.8

Otherwise, I will encounter “Transient resolver failure” error message.

Run command when gdb breakpoint is hit

I need to analyse a large pcap file and find problematic packets, so I want gdb automatically outputs packet index when error occurs. Below are the gdb commands:

(gdb) b packet.c:1430
Breakpoint 2 at 0x7ffff0f910aa: file packet.c, line 1430.
(gdb) commands
Type commands for breakpoint(s) 2, one per line.
End with a line saying just "end".
>silent
>frame 10
>printf "frame.number == %zu ||\n", packet_index
>c
>end
(gdb) r

Or you put them in a script:

b packet.c:1430
commands
 silent
 frame 10
 printf "frame.number == %zu ||\n", packet_index
 c
end

And it should be handy to log the output for checking later:

(gdb) set logging on

Use gdb’s convenience functions

Today I tried to set a conditional breakpoint in my program when a string variable is assigned one specific value:

b foo.c:488 if (int)strcmp(foo, "foo") == 0

But unfortunately, the gdb will exit with following error:

Unable to restore previously selected frame:
Selected thread is running.
terminate called after throwing an instance of 'gdb_exception_error'
Aborted

After checking in stackoverflow, I found gdb‘s convenience functions. So using $_streq instead of strcmp:

b foo.c:488 if $_streq(foo, "foo")

The gdb works like a charm!