On Thu, 2 Feb 2023, Uwe Kleine-König wrote:
> On Thu, Feb 02, 2023 at 01:26:08PM +0200, Ilpo Järvinen wrote:
> > On Thu, 2 Feb 2023, Uwe Kleine-König wrote:
> >
> > > The relevant history introducing serial_lsr_in() looks as follows:
> > >
> > > $ git log --graph --oneline --boundary 9fafe733514b..df36f3e3fbb7 -- drivers/tty/serial/8250/8250_port.c
> > > * df36f3e3fbb7 Merge tag 'v5.19-rc3' into tty-next
> > > |\
> > > | * be03b0651ffd serial: 8250: Store to lsr_save_flags after lsr read
> > > * | ...
> > > * | bdb70c424df1 serial: 8250: Create serial_lsr_in()
> > > * | ce338e4477cf serial: 8250: Store to lsr_save_flags after lsr read
> > > * | ...
> > > |/
> > > o 9fafe733514b tty: remove CMSPAR ifdefs
> > >
> > > So the patch "serial: 8250: Store to lsr_save_flags after lsr read" was
> > > introduced twice and in one branch it was followed up by commit
> > > bdb70c424df1 ("serial: 8250: Create serial_lsr_in()") which moved
> > > explicit lsr_saved_flags handling into a new function serial_lsr_in().
> > > When the two branches were merged in commit df36f3e3fbb7, we got both,
> > > serial_lsr_in() and the explicit lsr_saved_flags handling.
> > >
> > > So drop the explicit lsr_saved_flags handling.
> > >
> > > Fixes: df36f3e3fbb7 ("Merge tag 'v5.19-rc3' into tty-next")
> > > Signed-off-by: Uwe Kleine-König <u.kleine-koenig@xxxxxxxxxxxxxx>
> > > ---
> > > drivers/tty/serial/8250/8250_port.c | 2 --
> > > 1 file changed, 2 deletions(-)
> > >
> > > diff --git a/drivers/tty/serial/8250/8250_port.c b/drivers/tty/serial/8250/8250_port.c
> > > index 33be7aad11ef..e61753c295d5 100644
> > > --- a/drivers/tty/serial/8250/8250_port.c
> > > +++ b/drivers/tty/serial/8250/8250_port.c
> > > @@ -1512,8 +1512,6 @@ static inline void __stop_tx(struct uart_8250_port *p)
> > > u16 lsr = serial_lsr_in(p);
> > > u64 stop_delay = 0;
> > >
> > > - p->lsr_saved_flags |= lsr & LSR_SAVE_FLAGS;
> > > -
> > > if (!(lsr & UART_LSR_THRE))
> > > return;
> > > /*
> >
> > I don't know if Fixes tag is appropriate here. This fixes the mismerge
> > yes, however, the removed line itself seems harmless so there's no real
> > problem to fix.
>
> It might make a difference if LSR_SAVE_FLAGS != p->lsr_save_mask.
But currently lsr_save_mask always has at least LSR_SAVE_FLAGS bits so
that OR is no-op.
--
i.
