linuxppc-dev.lists.ozlabs.org archive mirror
 help / color / mirror / Atom feed
From: Benjamin Herrenschmidt <benh@kernel.crashing.org>
To: Gavin Shan <shangw@linux.vnet.ibm.com>
Cc: linuxppc-dev@lists.ozlabs.org
Subject: Re: [PATCH 23/27] powernv/opal: Notifier for OPAL events
Date: Wed, 12 Jun 2013 10:32:29 +1000	[thread overview]
Message-ID: <1370997149.8250.158.camel@pasglop> (raw)
In-Reply-To: <1370417668-16832-24-git-send-email-shangw@linux.vnet.ibm.com>

On Wed, 2013-06-05 at 15:34 +0800, Gavin Shan wrote:
> The patch intends to implement the notifier for variable OPAL events.
> It's notable that the notifier can be disabled dynamically. Also, the
> notifier could be fired upon incoming OPAL interrupts, or enabling
> the OPAL notifier.

"This patch implements a notifier to receive a notification on OPAL
event mask changes." is probably better. No need to blurb about
enable/disable, however add something along the lines of

"The notifier is only called as a result of an OPAL interrupt, which
will happen upon reception of FSP messages or PCI errors. Any event
mask change detected as a result of opal_poll_events() will not result
in a notifier call.

With OPALv3, opal_poll_event() will not clear interrupt conditions from
the FSP however, even if it consumes the messages (and thus updates the
event mask). Thus the interrupt notifier is a reliable way to get
the completion for FSP based OPAL operations. The specific list will
be added to the header file.


> Signed-off-by: Gavin Shan <shangw@linux.vnet.ibm.com>
> ---
>  arch/powerpc/include/asm/opal.h       |    3 +
>  arch/powerpc/platforms/powernv/opal.c |   79 ++++++++++++++++++++++++++++++++-
>  2 files changed, 81 insertions(+), 1 deletions(-)
> 
> diff --git a/arch/powerpc/include/asm/opal.h b/arch/powerpc/include/asm/opal.h
> index 2880797..64e7c84 100644
> --- a/arch/powerpc/include/asm/opal.h
> +++ b/arch/powerpc/include/asm/opal.h
> @@ -644,6 +644,9 @@ extern void hvc_opal_init_early(void);
>  extern int early_init_dt_scan_opal(unsigned long node, const char *uname,
>  				   int depth, void *data);
>  
> +extern int opal_notifier_register(uint64_t mask, void (*cb)(uint64_t));
> +extern void opal_notifier_enable(bool enable);

Make it two functions

opal_enable_notifier() vs. opal_disable_notifier()

>  extern int opal_get_chars(uint32_t vtermno, char *buf, int count);
>  extern int opal_put_chars(uint32_t vtermno, const char *buf, int total_len);
>  
> diff --git a/arch/powerpc/platforms/powernv/opal.c b/arch/powerpc/platforms/powernv/opal.c
> index 628c564..9bbbf93 100644
> --- a/arch/powerpc/platforms/powernv/opal.c
> +++ b/arch/powerpc/platforms/powernv/opal.c
> @@ -26,11 +26,20 @@ struct opal {
>  	u64 entry;
>  } opal;
>  
> +struct opal_cb {
> +	struct list_head list;
> +	uint64_t mask;
> +	void (*cb)(uint64_t);
> +};
> +
>  static struct device_node *opal_node;
>  static DEFINE_SPINLOCK(opal_write_lock);
>  extern u64 opal_mc_secondary_handler[];
>  static unsigned int *opal_irqs;
>  static unsigned int opal_irq_count;
> +static LIST_HEAD(opal_notifier);
> +static DEFINE_SPINLOCK(opal_notifier_lock);
> +static atomic_t opal_notifier_hold = ATOMIC_INIT(0);
>  
>  int __init early_init_dt_scan_opal(unsigned long node,
>  				   const char *uname, int depth, void *data)
> @@ -95,6 +104,74 @@ static int __init opal_register_exception_handlers(void)
>  
>  early_initcall(opal_register_exception_handlers);
>  
> +int opal_notifier_register(uint64_t mask, void (*cb)(uint64_t))
> +{
> +	unsigned long flags;
> +	struct opal_cb *p, *tmp;
> +
> +	if (!mask || !cb) {
> +		pr_warning("%s: Invalid argument (%llx, %p)!\n",
> +			__func__, mask, cb);
> +		return -EINVAL;
> +	}
> +
> +	p = kzalloc(sizeof(*p), GFP_KERNEL);
> +	if (!p) {
> +		pr_warning("%s: Out of memory (%llx, %p)!\n",
> +			__func__, mask, cb);
> +		return -ENOMEM;
> +	}
> +	p->mask = mask;
> +	p->cb   = cb;
> +
> +	spin_lock_irqsave(&opal_notifier_lock, flags);
> +	list_for_each_entry(tmp, &opal_notifier, list) {
> +		if (tmp->cb == cb || tmp->mask & mask) {
> +			pr_warning("%s: Duplicate evnet handler (%llx, %p)\n",
> +				__func__, tmp->mask, tmp->cb);
> +			spin_unlock_irqrestore(&opal_notifier_lock, flags);
> +			kfree(p);
> +			return -EEXIST;
> +		}
> +	}

Don't bother with checking the list already. This is not useful. Also
it's fine for two things to listen on the same event.

> +
> +	list_add_tail(&p->list, &opal_notifier);
> +	spin_unlock_irqrestore(&opal_notifier_lock, flags);
> +
> +	return 0;
> +}
> +
> +static void opal_do_notifier(uint64_t events)
> +{
> +	struct opal_cb *tmp;
> +
> +	if (atomic_read(&opal_notifier_hold))
> +		return;
> +	if (!events)
> +		return;
> +
> +	list_for_each_entry(tmp, &opal_notifier, list) {
> +		if (events & tmp->mask)
> +			tmp->cb(events & tmp->mask);
> +	}
> +}

My idea was to call this if the event bit has changed since the last
time we called opal_do_notifier. IE. Use a static last_notified_mask
and do something like

	changed_mask = last_notified_mask ^ events;

	list_for_each_entry(tmp, &opal_notifier, list) {
		if (changed_mask & tmp->mask)
			tmp->cb(events);

Also, always pass the whole events to the callback, no point in
filtering.

BTW, "tmp" isn't a nice name here.

> +void opal_notifier_enable(bool enable)
> +{
> +	int64_t rc;
> +	uint64_t evt = 0;
> +
> +	if (enable) {
> +		atomic_set(&opal_notifier_hold, 0);
> +
> +		/* Process pending events */
> +		rc = opal_poll_events(&evt);
> +		if (rc == OPAL_SUCCESS && evt)
> +			opal_do_notifier(evt);
> +	} else
> +		atomic_set(&opal_notifier_hold, 1);
> +}

As I said, two functions.

>  int opal_get_chars(uint32_t vtermno, char *buf, int count)
>  {
>  	s64 len, rc;
> @@ -297,7 +374,7 @@ static irqreturn_t opal_interrupt(int irq, void *data)
>  
>  	opal_handle_interrupt(virq_to_hw(irq), &events);
>  
> -	/* XXX TODO: Do something with the events */
> +	opal_do_notifier(events);
>  
>  	return IRQ_HANDLED;
>  }

Cheers,
Ben.

  reply	other threads:[~2013-06-12  0:32 UTC|newest]

Thread overview: 43+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2013-06-05  7:34 [PATCH v3 00/27] EEH Support for PowerNV platform Gavin Shan
2013-06-05  7:34 ` [PATCH 01/27] powerpc/eeh: Fix fetching bus for single-dev-PE Gavin Shan
2013-06-05  7:34 ` [PATCH 02/27] powerpc/eeh: Enhance converting EEH dev Gavin Shan
2013-06-05  7:34 ` [PATCH 03/27] powerpc/eeh: Make eeh_phb_pe_get() public Gavin Shan
2013-06-05  7:34 ` [PATCH 04/27] powerpc/eeh: Make eeh_pe_get() public Gavin Shan
2013-06-05  7:34 ` [PATCH 05/27] powerpc/eeh: Trace PCI bus from PE Gavin Shan
2013-06-05  7:34 ` [PATCH 06/27] powerpc/eeh: Make eeh_init() public Gavin Shan
2013-06-05  7:34 ` [PATCH 07/27] powerpc/eeh: EEH post initialization operation Gavin Shan
2013-06-05  7:34 ` [PATCH 08/27] powerpc/eeh: Refactor eeh_reset_pe_once() Gavin Shan
2013-06-05  7:34 ` [PATCH 09/27] powerpc/eeh: Delay EEH probe during hotplug Gavin Shan
2013-06-05  7:34 ` [PATCH 10/27] powerpc/eeh: Differentiate EEH events Gavin Shan
2013-06-05  7:34 ` [PATCH 11/27] powerpc/eeh: Sync OPAL API with firmware Gavin Shan
2013-06-05  7:34 ` [PATCH 12/27] powerpc/eeh: EEH backend for P7IOC Gavin Shan
2013-06-05  7:34 ` [PATCH 13/27] powerpc/eeh: I/O chip post initialization Gavin Shan
2013-06-05  7:34 ` [PATCH 14/27] powerpc/eeh: I/O chip EEH enable option Gavin Shan
2013-06-05  7:34 ` [PATCH 15/27] powerpc/eeh: I/O chip EEH state retrieval Gavin Shan
2013-06-11  7:37   ` Benjamin Herrenschmidt
2013-06-12  3:32     ` Gavin Shan
2013-06-12  4:19       ` Benjamin Herrenschmidt
2013-06-13  4:26         ` Gavin Shan
2013-06-13  4:42           ` Benjamin Herrenschmidt
2013-06-13  5:50             ` Gavin Shan
2013-06-05  7:34 ` [PATCH 16/27] powerpc/eeh: I/O chip PE reset Gavin Shan
2013-06-05  7:34 ` [PATCH 17/27] powerpc/eeh: I/O chip PE log and bridge setup Gavin Shan
2013-06-11  7:37   ` Benjamin Herrenschmidt
2013-06-12  3:33     ` Gavin Shan
2013-06-05  7:34 ` [PATCH 18/27] powerpc/eeh: PowerNV EEH backends Gavin Shan
2013-06-05  7:34 ` [PATCH 19/27] powerpc/eeh: Initialization for PowerNV Gavin Shan
2013-06-05  7:34 ` [PATCH 20/27] powerpc/eeh: Enable EEH check for config access Gavin Shan
2013-06-05  7:34 ` [PATCH 21/27] powerpc/eeh: Process interrupts caused by EEH Gavin Shan
2013-06-11  8:13   ` Benjamin Herrenschmidt
2013-06-13  4:14     ` Gavin Shan
2013-06-05  7:34 ` [PATCH 22/27] powerpc/eeh: Allow to check fenced PHB proactively Gavin Shan
2013-06-05  7:34 ` [PATCH 23/27] powernv/opal: Notifier for OPAL events Gavin Shan
2013-06-12  0:32   ` Benjamin Herrenschmidt [this message]
2013-06-12  3:15     ` Gavin Shan
2013-06-05  7:34 ` [PATCH 24/27] powernv/opal: Disable OPAL notifier upon poweroff Gavin Shan
2013-06-05  7:34 ` [PATCH 25/27] powerpc/eeh: Register OPAL notifier for PCI error Gavin Shan
2013-06-05  7:34 ` [PATCH 26/27] powerpc/powernv: Debugfs directory for PHB Gavin Shan
2013-06-05  7:34 ` [PATCH 27/27] powerpc/eeh: Debugfs for error injection Gavin Shan
2013-06-11  7:46 ` [PATCH v3 00/27] EEH Support for PowerNV platform Benjamin Herrenschmidt
2013-06-12  3:18   ` Gavin Shan
  -- strict thread matches above, loose matches on Subject: below --
2013-06-15  9:02 [PATCH v4 " Gavin Shan
2013-06-15  9:03 ` [PATCH 23/27] powernv/opal: Notifier for OPAL events Gavin Shan

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=1370997149.8250.158.camel@pasglop \
    --to=benh@kernel.crashing.org \
    --cc=linuxppc-dev@lists.ozlabs.org \
    --cc=shangw@linux.vnet.ibm.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).