On Wed, 23 Jul 2003, Daniel Ritz wrote:
> [shortening the cc: list a bit..]
>
> On Wed July 23 2003 12:26, Javier Achirica wrote:
> >
> > You cannot use down() in xmit, as it may be called in interrupt context. I
> > know it slows things down, but that's the only way I figured out of
> > handling a transmission while the card is processing a long command.
>
> hu? no. you can do a down() as xmit is never called from interrupt context. and
> the dev->hard_start_xmit() calls are serialized with the dev->xmit_lock. the
> serialization is broken by the schedule_work() thing.
Ok, sure. But then, if you are in interrupt context, whet do you do with
the packet?
There's no problem with serialization, because if the packet gets
scheduled with schedule_work(), the queue it's also stopped (before
releasing the xmit_lock) so dev->hard_start_xmit() doesn't get called
again.
> >
> > I thought about the fix and I think it's fixed. The only case the race
> > could happen is if there's some work pending to be scheduled and the queue
> > gets started again (by the interrupt handler), so airo_start_xmit
> > overwrites the priv->xmit data. Now, because of the new flag, the
> > interrupt handler won't wake the queue until the pending packet is
> > sent to the card (or fails) so I don't see how can the race happen
> > (although I didn't see it until you pointed out :-(
> >
>
> may be the flag fixes the problem, but it adds complexity...
I know. It's the best I could do. Anyway, in the works there's a complete
rewrite of the packet handling, moving from I/O port access to mapped
memory. I'll get into that once I get this version stable and the other
one working.
Javier Achirica
> >
> > On Wed, 23 Jul 2003, Daniel Ritz wrote:
> >
> > > ok, now the braindamaged thing called sourceforge showed the changes, but:
> > > - i don't think the race is fixed. just remove the whole down_trylock()
> > > crap in the xmit altogether and replace it with a single down(). faster,
> > > simpler, not racy...and with the schedule_work you win nothing, you lose
> > > speed
> > > - please don't commit bugfixes and new features in the same changeset...
> > > - the loop-forever fix in transmit_allocate: you should have copied the
> > > comment
> > > changes from my patch too, so the spin-forever-comment goes away...
> > >
> > > i look closer when i'm home, having a real operating system to work on, not
> > > this
> > > winblows box at work now..
> > >
> > > -daniel
> > >
> > >
> > > Javier Achirica wrote:
> > > >
> > > > Today I updated the CVS and Sourceforge (airo-linux.sf.net) with the
> > > > latest version (1.53) that (I hope) fixes the race problem. If everything
> > > > is fine, I'll commit the changes to the kernel tree.
> > > >
> > > > Javier Achirica
> > > >
> > > > On Mon, 21 Jul 2003, Daniel Ritz wrote:
> > > >
> > > > > On Mon July 21 2003 21:44, Javier Achirica wrote:
> > > > > >
> > > > > > On Mon, 21 Jul 2003, Daniel Ritz wrote:
> > > > > >
> > > > > > > On Mon July 21 2003 13:00, Javier Achirica wrote:
> > > > > > > >
> > > > > > > > Daniel,
> > > > > > > >
> > > > > > > > Thank you for your patch. Some comments about it:
> > > > > > > >
> > > > > > > > - I'd rather fix whatever is broken in the current code than going
> > > back to
> > > > > > > > spinlocks, as they increase latency and reduce concurrency. In any
> > > case,
> > > > > > > > please check your code. I've seen a spinlock in the interrupt
> > > handler that
> > > > > > > > may lock the system.
> > > > > > >
> > > > > > > but we need to protect from interrupts while accessing the card and
> > > waiting for
> > > > > > > completion. semaphores don't protect you from that.
> > > spin_lock_irqsave does. the
> > > > > > > spin_lock in the interrupt handler is there to protect from
> > > interrupts from
> > > > > > > other processors in a SMP system (see Documentation/spinlocks.txt)
> > > and is btw.
> > > > > > > a no-op on UP. and semaphores are quite heavy....
> > > > > >
> > > > > > Not really. You can still read the received packets from the card (as
> > > > > > you're not issuing any command and are using the other BAP) while a
> > > > > > command is in progress. There are some specific cases in which you
> > > need
> > > > > > to have protection, and that cases are avoided with the down_trylock.
> > > > > >
> > > > >
> > > > > ok, i think i have to look closer...if the card can handle that then we
> > > don't need
> > > > > to irq-protect all the areas i did protect...but i do think that those
> > > down_trylock and
> > > > > then the schedule_work should be replaced by a simple
> > > spinlock_irq_save...
> > > > >
> > > > > i look closer at it tomorrow.
> > > > > you happen to have the tech spec lying aroung?
> > > > >
> > > > > > AFAIK, interrupt serialization is assured by the interrupt handler, so
> > > you
> > > > > > don't need to do that.
> > > > > >
> > > > > > > > - The fix for the transmit code you mention, is about fixing the
> > > returned
> > > > > > > > value in case of error? If not, please explain it to me as I don't
> > > see any
> > > > > > > > other changes.
> > > > > > >
> > > > > > > fixes:
> > > > > > > - return values
> > > > > > > - when to free the skb, when not
> > > > > > > - disabling the queues
> > > > > > > - netif_wake_queue called from the interrupt handler only (and on
> > > the right
> > > > > > > net_device)
> > > > > > > - i think the priv->xmit stuff and then the schedule_work is evil:
> > > > > > > if you return 0 from the dev->hard_start_xmit then the network
> > > layer assumes
> > > > > > > that the packet was kfree_skb()'ed (which does only frees the
> > > packet when the
> > > > > > > refcount drops to zero.) this is the cause for the keventd
> > > killing, for sure!
> > > > > > >
> > > > > > > if you return 0 you already kfree_skb()'ed the packet. and that's
> > > it.
> > > > > >
> > > > > > This is where I have the biggest problems. As I've read in
> > > > > > Documentation/networking/driver.txt, looks like the packet needs to be
> > > > > > freed "soon", but doesn't require to be before returning 0 in
> > > > > > hard_start_xmit. Did I get it wrong?
> > > > > >
> > > > >
> > > > > no, i got it wrong. but still...it's the xmit where the oops comes
> > > from....
> > > > >
> > > > > wait. isn't there a race in airo_do_xmit? at high xfer rates (when it
> > > oopses) the
> > > > > queue can wake right after it is stopped in the down_trylock section. so
> > > you can
> > > > > happen to loose an skb 'cos the write to priv->xmit is not protected at
> > > all and
> > > > > there should be a check so that only one skb can be queue there. no?
> > > > > (and then the irq-handler can wake the queue too)
> > > > >
> > > > > ok, i think i got it now. i'll do a new patch tomorrow or so that tries:
> > > > > - to fix the transmit not to oops
> > > > > - to avoid disabling the irq's whenever possible
> > > > > - using spinlocks instead of the heavier semaphores ('cos i think if
> > > it's done cleaner
> > > > > than i did it now, it's faster than the semas, and to make hch happy
> > > :)
> > > > >
> > > > >
> > > > > > Thanks for your help,
> > > > > > Javier Achirica
> > > > > >
> > > > >
> > > > > rgds
> > > > > -daniel
> > > > >
> > > > >
> > > > >
> > > >
> > >
> > >
> > >
> >
> >
>
>
>
-
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majordomo@vger.kernel.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/
This archive was generated by hypermail 2b29 : Wed Jul 23 2003 - 22:00:50 EST