On (10/16/18 14:54), Peter Zijlstra wrote:
>
> No, no wakups. irq_work to wake the printk-thread, at most.
>
There are cases when we probably prefer to be in "direct printk" mode.
E.g. sysrq or late PM stages (probably).
Doing irq_work->wake_up_process->printk_kthread from sysrq probably
might
On Wed 2018-10-17 16:00:44, Peter Zijlstra wrote:
> On Wed, Oct 17, 2018 at 12:50:15PM +0200, Petr Mladek wrote:
> > Also note that by deferred printk I mean deferring the console
> > handling! IMHO, there are _no more problems_ with storing
> > the messages into the buffer if we accept that the cu
On Wed, Oct 17, 2018 at 12:50:15PM +0200, Petr Mladek wrote:
> > If you have a lockless buffer and a trailing printk thread, that's
> > deferred.
> >
> > And earlycon _should_ be lockless (yes, I know, some suck)
> >
> > But if you do this deferred nonsense that's currently in printk, then
> > y
On (10/17/18 09:57), Peter Zijlstra wrote:
> On Wed, Oct 17, 2018 at 01:32:51PM +0900, Sergey Senozhatsky wrote:
> > This probably will be a bit more hairy. logbuf is written to by many
> > sources and is read from by many sides, including user-space [both read()
> > and write()]. So we will need m
On Tue 2018-10-16 14:17:52, Peter Zijlstra wrote:
> On Tue, Oct 16, 2018 at 01:40:06PM +0200, Petr Mladek wrote:
> > On Tue 2018-10-16 09:27:19, Peter Zijlstra wrote:
>
> > > Instead of this tinkering around the edges, why don't you make the main
> > > logbuf a lockless ringbuffer and then delegat
On Wed, Oct 17, 2018 at 01:32:51PM +0900, Sergey Senozhatsky wrote:
> This probably will be a bit more hairy. logbuf is written to by many
> sources and is read from by many sides, including user-space [both read()
> and write()]. So we will need more flags/magic around memcpy(). A simple,
> "grab
On (10/16/18 14:54), Peter Zijlstra wrote:
> On Tue, Oct 16, 2018 at 09:27:34PM +0900, Sergey Senozhatsky wrote:
> > per-CPU printk_safe _semi-magic_ makes some things simple to handle.
> > We can't just remove per-CPU buffers and add a wake_up_process() at
> > the bottom of vprintk_emit(). Because
On Tue, Oct 16, 2018 at 02:54:15PM +0200, Peter Zijlstra wrote:
> printk will determine the current context:
>
> task, softirq, hardirq or NMI
>
> and pick the corresponding per-cpu line buffer and do the vsnprintf()
We need 4, but we don't need to do the exact context determination for
this.
On Tue, Oct 16, 2018 at 09:27:34PM +0900, Sergey Senozhatsky wrote:
> per-CPU printk_safe _semi-magic_ makes some things simple to handle.
> We can't just remove per-CPU buffers and add a wake_up_process() at
> the bottom of vprintk_emit(). Because this will deadlock:
>
> printk()
>wake_up_p
On Tue, Oct 16, 2018 at 09:27:34PM +0900, Sergey Senozhatsky wrote:
> On (10/16/18 09:27), Peter Zijlstra wrote:
> >
> > So I really _really_ hate all this. Utterly detest it.
>
> I learned a new word today - detest. I can haz re-entrant consoles now please?
early_serial_write() not haz a lock a
On (10/16/18 09:27), Peter Zijlstra wrote:
>
> So I really _really_ hate all this. Utterly detest it.
I learned a new word today - detest. I can haz re-entrant consoles now please?
> That whole magic 'safe' printk buffer crap is just that crap.
No, I don't see it this way; printk_safe is *semi-
On Tue, Oct 16, 2018 at 01:40:06PM +0200, Petr Mladek wrote:
> On Tue 2018-10-16 09:27:19, Peter Zijlstra wrote:
> > Instead of this tinkering around the edges, why don't you make the main
> > logbuf a lockless ringbuffer and then delegate the actual printing of
> > that buffer to a kthread, excep
On Tue 2018-10-16 09:27:19, Peter Zijlstra wrote:
> On Tue, Oct 16, 2018 at 02:04:26PM +0900, Sergey Senozhatsky wrote:
> > Make printk_safe_enter_irqsave()/etc macros available to the
> > rest of the kernel.
> >
> > Signed-off-by: Sergey Senozhatsky
> > ---
> > include/linux/printk.h | 40
On Tue, Oct 16, 2018 at 02:04:26PM +0900, Sergey Senozhatsky wrote:
> Make printk_safe_enter_irqsave()/etc macros available to the
> rest of the kernel.
>
> Signed-off-by: Sergey Senozhatsky
> ---
> include/linux/printk.h | 40 +
> kernel/printk/internal.
Make printk_safe_enter_irqsave()/etc macros available to the
rest of the kernel.
Signed-off-by: Sergey Senozhatsky
---
include/linux/printk.h | 40 +
kernel/printk/internal.h| 37 --
kernel/printk/printk_safe.c | 6 +++
15 matches
Mail list logo
