strcpy() performs no bounds checking on the destination buffer. This
could result in linear overflows beyond the end of the buffer, leading
to all kinds of misbehaviors. The safe replacement is strscpy() but in
this case it is simpler to add NULL to the first position since we want
to empty the string.
This is a previous step in the path to remove the strcpy() function.
Signed-off-by: Len Baker <[email protected]>
---
drivers/staging/fbtft/fbtft-core.c | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/drivers/staging/fbtft/fbtft-core.c b/drivers/staging/fbtft/fbtft-core.c
index 3723269890d5..b8791806cb20 100644
--- a/drivers/staging/fbtft/fbtft-core.c
+++ b/drivers/staging/fbtft/fbtft-core.c
@@ -1037,7 +1037,7 @@ int fbtft_init_display(struct fbtft_par *par)
case -1:
i++;
/* make debug message */
- strcpy(msg, "");
+ msg[0] = 0;
j = i + 1;
while (par->init_sequence[j] >= 0) {
sprintf(str, "0x%02X ", par->init_sequence[j]);
--
2.25.1
On Sun, Jul 18, 2021 at 4:43 PM Len Baker <[email protected]> wrote:
>
> strcpy() performs no bounds checking on the destination buffer. This
> could result in linear overflows beyond the end of the buffer, leading
> to all kinds of misbehaviors. The safe replacement is strscpy() but in
> this case it is simpler to add NULL to the first position since we want
> to empty the string.
> This is a previous step in the path to remove the strcpy() function.
Any document behind this (something to read on the site(s) more or
less affiliated with what is going to happen in the kernel) to read
background?
...
> case -1:
> i++;
> /* make debug message */
> - strcpy(msg, "");
> + msg[0] = 0;
Strictly speaking it should be '\0'.
> j = i + 1;
> while (par->init_sequence[j] >= 0) {
> sprintf(str, "0x%02X ", par->init_sequence[j]);
--
With Best Regards,
Andy Shevchenko
On Sun, Jul 18, 2021 at 9:43 PM Andy Shevchenko
<[email protected]> wrote:
> On Sun, Jul 18, 2021 at 4:43 PM Len Baker <[email protected]> wrote:
> > strcpy() performs no bounds checking on the destination buffer. This
> > could result in linear overflows beyond the end of the buffer, leading
> > to all kinds of misbehaviors. The safe replacement is strscpy() but in
> > this case it is simpler to add NULL to the first position since we want
"NULL" is a pointer value, "NUL" is the character with value zero.
> > to empty the string.
>
> > This is a previous step in the path to remove the strcpy() function.
>
> Any document behind this (something to read on the site(s) more or
> less affiliated with what is going to happen in the kernel) to read
> background?
>
> ...
>
> > case -1:
> > i++;
> > /* make debug message */
> > - strcpy(msg, "");
While this strcpy() is provably safe at compile-time, and will probably
be replaced by an assignment to zero by the compiler...
> > + msg[0] = 0;
>
> Strictly speaking it should be '\0'.
>
> > j = i + 1;
> > while (par->init_sequence[j] >= 0) {
> > sprintf(str, "0x%02X ", par->init_sequence[j]);
... the real danger is the
strcat(msg, str);
on the next line.
Fortunately this whole debug printing block (including the strcpy)
can (and should) be rewritten to just use "%*ph".
Gr{oetje,eeting}s,
Geert
--
Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]
In personal conversations with technical people, I call myself a hacker. But
when I'm talking to journalists I just say "programmer" or something like that.
-- Linus Torvalds
Hi,
On Sun, Jul 18, 2021 at 10:42:42PM +0300, Andy Shevchenko wrote:
> On Sun, Jul 18, 2021 at 4:43 PM Len Baker <[email protected]> wrote:
> >
> > strcpy() performs no bounds checking on the destination buffer. This
> > could result in linear overflows beyond the end of the buffer, leading
> > to all kinds of misbehaviors. The safe replacement is strscpy() but in
> > this case it is simpler to add NULL to the first position since we want
> > to empty the string.
>
> > This is a previous step in the path to remove the strcpy() function.
>
> Any document behind this (something to read on the site(s) more or
> less affiliated with what is going to happen in the kernel) to read
> background?
This is a task of the KSPP (kernel self protection project) [1]
[1] https://github.com/KSPP/linux/issues/88
>
> ...
>
> > case -1:
> > i++;
> > /* make debug message */
> > - strcpy(msg, "");
> > + msg[0] = 0;
>
> Strictly speaking it should be '\0'.
Ok, understood.
>
> > j = i + 1;
> > while (par->init_sequence[j] >= 0) {
> > sprintf(str, "0x%02X ", par->init_sequence[j]);
>
>
> --
> With Best Regards,
> Andy Shevchenko
Thanks for the feedback,
Len
On Mon, Jul 19, 2021 at 09:53:29AM +0200, Geert Uytterhoeven wrote:
> On Sun, Jul 18, 2021 at 9:43 PM Andy Shevchenko
> <[email protected]> wrote:
> > On Sun, Jul 18, 2021 at 4:43 PM Len Baker <[email protected]> wrote:
> > > strcpy() performs no bounds checking on the destination buffer. This
> > > could result in linear overflows beyond the end of the buffer, leading
> > > to all kinds of misbehaviors. The safe replacement is strscpy() but in
> > > this case it is simpler to add NULL to the first position since we want
>
> "NULL" is a pointer value, "NUL" is the character with value zero.
Ok, understood. Thanks.
>
> > > to empty the string.
> >
> > > This is a previous step in the path to remove the strcpy() function.
> >
> > Any document behind this (something to read on the site(s) more or
> > less affiliated with what is going to happen in the kernel) to read
> > background?
> >
> > ...
> >
> > > case -1:
> > > i++;
> > > /* make debug message */
> > > - strcpy(msg, "");
>
> While this strcpy() is provably safe at compile-time, and will probably
> be replaced by an assignment to zero by the compiler...
>
> > > + msg[0] = 0;
> >
> > Strictly speaking it should be '\0'.
> >
> > > j = i + 1;
> > > while (par->init_sequence[j] >= 0) {
> > > sprintf(str, "0x%02X ", par->init_sequence[j]);
>
> ... the real danger is the
>
> strcat(msg, str);
>
> on the next line.
> Fortunately this whole debug printing block (including the strcpy)
> can (and should) be rewritten to just use "%*ph".
Ok, I will work on it and I will send a v2 for review. Thanks for the
feedback.
>
> Gr{oetje,eeting}s,
>
> Geert
>
> --
> Geert Uytterhoeven -- There's lots of Linux beyond ia32 -- [email protected]
>
> In personal conversations with technical people, I call myself a hacker. But
> when I'm talking to journalists I just say "programmer" or something like that.
> -- Linus Torvalds
Regards,
Len