As well known, hvc backend can register its opertions to hvc backend.
the opertions contain put_chars(), get_chars() and so on.
Some hvc backend may do dma in its opertions. eg, put_chars() of
virtio-console. But in the code of hvc framework, it may pass DMA
incapable memory to put_chars() under a specific configuration, which
is explained in commit c4baad5029(virtio-console: avoid DMA from stack):
1, c[] is on stack,
hvc_console_print():
char c[N_OUTBUF] __ALIGNED__;
cons_ops[index]->put_chars(vtermnos[index], c, i);
2, ch is on stack,
static void hvc_poll_put_char(,,char ch)
{
struct tty_struct *tty = driver->ttys[0];
struct hvc_struct *hp = tty->driver_data;
int n;
do {
n = hp->ops->put_chars(hp->vtermno, &ch, 1);
} while (n <= 0);
}
Commit c4baad5029 is just the fix to avoid DMA from stack memory, which
is passed to virtio-console by hvc framework in above code. But I think
the fix is aggressive, it directly uses kmemdup() to alloc new buffer
from kmalloc area and do memcpy no matter the memory is in kmalloc area
or not. But most importantly, it should better be fixed in the hvc
framework, by changing it to never pass stack memory to the put_chars()
function in the first place. Otherwise, we still face the smae issue if
a new hvc backend using dma added in the furture.
Considering lock competition of hp->outbuf and the complicated logic in
hvc_console_print(), I didn’t use hp->outbuf, just allocate additional
memory(length N_OUTBUF) and append it to hp->outbuf.
For the issue in hvc_poll_put_char(), I use a static char to replace
the char in stack.
With the patch, we can remove the fix c4baad5029.
Signed-off-by: Xianting Tian <redacted>
Tested-by: Xianting Tian <redacted>
---
drivers/tty/hvc/hvc_console.c | 14 +++++++++++---
1 file changed, 11 insertions(+), 3 deletions(-)
On Sun, Aug 1, 2021 at 7:16 AM Xianting Tian
[off-list ref] wrote:
Considering lock competition of hp->outbuf and the complicated logic in
hvc_console_print(), I didn’t use hp->outbuf, just allocate additional
memory(length N_OUTBUF) and append it to hp->outbuf.
For the issue in hvc_poll_put_char(), I use a static char to replace
the char in stack.
While this may work, it sounds rather obscure to me, I don't think
it's a good idea
to append the buffer at the back.
If you need a separate field besides hp->outbuf, I would make that part of the
structure itself, and give it the correct alignment constraints to ensure it is
in a cache line by itself. The size of this field is a compile-time
constant, so I
don't see a need to play tricks with pointer arithmetic.
I'm not sure about the locking either: Is it possible for two CPUs to enter
hvc_console_print() at the same time, or is there locking at a higher level
already? It would be good to document this in the structure definition next
to the field.
quoted hunk
@@ -878,6 +885,7 @@ static void hvc_poll_put_char(struct tty_driver *driver, int line, char ch) struct tty_struct *tty = driver->ttys[0]; struct hvc_struct *hp = tty->driver_data; int n;+ static char ch = ch; do { n = hp->ops->put_chars(hp->vtermno, &ch, 1);
This does not compile, and it's neither thread-safe nor dma safe if you get it
to build by renaming the variable.
Arnd