Messages in this thread | | | From | Kees Cook <> | Subject | [PATCH 0/2] vsprintf: ignore %n again | Date | Mon, 16 Sep 2013 00:43:55 -0700 |
| |
Whether seq_printf should return void or error, %n still needs to be removed. As such, instead of changing the seq_file structure and adding instructions to all callers of seq_printf, just examine seq->count for the callers that care about how many characters were put into the buffer, as suggested by George Spelvin. First patch removes all %n usage in favor of checking seq->count before/after. Second patch makes %n ignore its argument.
Testing shows this all works happily, and everything is still getting padded correctly:
/proc/consoles: ttyS0 -W- (EC a) 4:64 netcon0 -W- (E )
/proc/self/maps: ... 01ee7000-01f08000 rw-p 00000000 00:00 0 [heap] 7fdc79bd4000-7fdc79bf6000 r-xp 00000000 fd:01 394247 /lib/x86_64-linux-gnu/libtinfo.so.5.9 ...
/proc/net/tcp sl local_address rem_address st tx_queue rx_queue tr tm->when retrnsmt uid timeout inode 0: 00000000:0016 00000000:0000 0A 00000000:00000000 00:00000000 00000000 0 0 2239 1 ffff88007bfd0000 100 0 0 10 0 ...
/proc/net/udp sl local_address rem_address st tx_queue rx_queue tr tm->when retrnsmt uid timeout inode ref pointer drops 12: 0DAAA8C0:D9D1 0100000A:0035 01 00000000:00000000 00:00000000 00000000 0 0 7534 2 ffff880078048000 0
And a test with a %n in a format string shows the warning:
[ 10.693638] ------------[ cut here ]------------ [ 10.693657] WARNING: CPU: 0 PID: 2048 at lib/vsprintf.c:1693 vsnprintf+0x5c1/0x600() [ 10.693660] Please remove ignored %n in '%n [ 10.693663] ' ...
Fixing the other callers of seq_printf to do the right thing (void or not) can be separate from this series.
-Kees
| |