Skip to contentSkip to conversation

mac80211: add general rate information to Tx status

RFC In review Tomas Winkler 5 participants 11 messages Feb 5, 2008 – Feb 7, 2008 mac80211

lore
Share & tools
lore
#0 lore raw

RFC v2 1/1 mac80211: add general rate information to Tx status

From: Ron Rindjunsky <[email protected]>

This patch introduces data structures meant to contain actual configuration in which Tx was made in HW, for general purposes and for rate scaling in specific.
it includes the rate, general flags for the rate (e.g. Tx antenna) and the number of retries for the rate.

---
 include/net/mac80211.h |   22 +++++++++++++++++++++-
 1 files changed, 21 insertions(+), 1 deletions(-)
include/net/mac80211.h +21 -1
index 460da54..ec966c8 100644
@@ -318,6 +318,22 @@ struct ieee80211_rx_status {
 };
 
 /**
+ * struct ieee80211_tx_rate_ctrl - rate ctrl data
+ *
+ * The low-level driver should provide this struct as part
+ * of the ieee80211_tx_status. this struct should give information
+ * about the tx that is not part of the plcp rates table.
+ * @bitrate: rate used.
+ * @flags: Tx information (e.g. antenna selectd, guard interval, etc.)
+ * @retry_count: number of retries for this specific rate
+ */
+struct ieee80211_tx_rate_ctrl {
+	u32 bitrate;
+	u32 flags;
+	u8 retry_count;
+};
+
+/**
  * enum ieee80211_tx_status_flags - transmit status flags
  *
  * Status flags to indicate various transmit conditions.
@@ -343,7 +359,7 @@ enum ieee80211_tx_status_flags {
  * @control: a copy of the &struct ieee80211_tx_control passed to the driver
  *	in the tx() callback.
  * @flags: transmit status flags, defined above
- * @retry_count: number of retries
+ * @retry_count: total number of retries
  * @excessive_retries: set to 1 if the frame was retried many times
  *	but not acknowledged
  * @ampdu_ack_len: number of aggregated frames.
@@ -353,6 +369,8 @@ enum ieee80211_tx_status_flags {
  * @ack_signal: signal strength of the ACK frame
  * @queue_length: ?? REMOVE
  * @queue_number: ?? REMOVE
+ * @rate_ctrl_num: number of rate_ctrl allocated
+ * @rate_ctrl: info for rate at which the packet was actually transmitted
  */
 struct ieee80211_tx_status {
 	struct ieee80211_tx_control control;
@@ -364,6 +382,8 @@ struct ieee80211_tx_status {
 	int ack_signal;
 	int queue_length;
 	int queue_number;
+	size_t rate_ctrl_num;
+	struct ieee80211_tx_rate_ctrl rate_ctrl[0];
 };
 
 /**
Signature · 9 lines
1.5.2.2

---------------------------------------------------------------------
Intel Israel (74) Limited

This e-mail and any attachments may contain confidential material for
the sole use of the intended recipient(s). Any review or distribution
by others is strictly prohibited. If you are not the intended
recipient, please contact the sender and delete all copies.
#1 lore raw
Quoted 69 lines · Tomas Winkler

From: Ron Rindjunsky <[email protected]>

This patch introduces data structures meant to contain actual configuration in which Tx was made in HW, for general purposes and for rate scaling in specific.
it includes the rate, general flags for the rate (e.g. Tx antenna) and the number of retries for the rate.

---
 include/net/mac80211.h |   22 +++++++++++++++++++++-
 1 files changed, 21 insertions(+), 1 deletions(-)
include/net/mac80211.h +21 -2
index 460da54..ec966c8 100644
@@ -318,6 +318,22 @@ struct ieee80211_rx_status {
 };
 
 /**
+ * struct ieee80211_tx_rate_ctrl - rate ctrl data
+ *
+ * The low-level driver should provide this struct as part
+ * of the ieee80211_tx_status. this struct should give information
+ * about the tx that is not part of the plcp rates table.
+ * @bitrate: rate used.
+ * @flags: Tx information (e.g. antenna selectd, guard interval, etc.)
+ * @retry_count: number of retries for this specific rate
+ */
+struct ieee80211_tx_rate_ctrl {
+       u32 bitrate;
+       u32 flags;
+       u8 retry_count;
+};
+
+/**
  * enum ieee80211_tx_status_flags - transmit status flags
  *
  * Status flags to indicate various transmit conditions.
@@ -343,7 +359,7 @@ enum ieee80211_tx_status_flags {
  * @control: a copy of the &struct ieee80211_tx_control passed to the driver
  *     in the tx() callback.
  * @flags: transmit status flags, defined above
- * @retry_count: number of retries
+ * @retry_count: total number of retries
  * @excessive_retries: set to 1 if the frame was retried many times
  *     but not acknowledged
  * @ampdu_ack_len: number of aggregated frames.
@@ -353,6 +369,8 @@ enum ieee80211_tx_status_flags {
  * @ack_signal: signal strength of the ACK frame
  * @queue_length: ?? REMOVE
  * @queue_number: ?? REMOVE
+ * @rate_ctrl_num: number of rate_ctrl allocated
+ * @rate_ctrl: info for rate at which the packet was actually transmitted
  */
 struct ieee80211_tx_status {
        struct ieee80211_tx_control control;
@@ -364,6 +382,8 @@ struct ieee80211_tx_status {
        int ack_signal;
        int queue_length;
        int queue_number;
+       size_t rate_ctrl_num;
+       struct ieee80211_tx_rate_ctrl rate_ctrl[0];
 };
 
 /**
--
Quoted 1 line

1.5.2.2

Felix will this work for you?

Thanks
Tomas

Quoted 13 lines

---------------------------------------------------------------------
Intel Israel (74) Limited

This e-mail and any attachments may contain confidential material for the sole use of the intended recipient(s). Any review or distribution by others is strictly prohibited. If you are not the intended recipient, please contact the sender and delete all copies.

-
To unsubscribe from this list: send the line "unsubscribe linux-wireless" in the body of a message to [email protected]
More majordomo info at http://vger.kernel.org/majordomo-info.html

#2 lore raw
Quoted 68 lines · Tomas Winkler
Quoted 67 lines · Tomas Winkler

From: Ron Rindjunsky <[email protected]>

This patch introduces data structures meant to contain actual configuration in which Tx was made in HW, for general purposes and for rate scaling in specific.
it includes the rate, general flags for the rate (e.g. Tx antenna) and the number of retries for the rate.

---
 include/net/mac80211.h |   22 +++++++++++++++++++++-
 1 files changed, 21 insertions(+), 1 deletions(-)
include/net/mac80211.h +21 -1
index 460da54..ec966c8 100644
@@ -318,6 +318,22 @@ struct ieee80211_rx_status {
 };
 
 /**
+ * struct ieee80211_tx_rate_ctrl - rate ctrl data
+ *
+ * The low-level driver should provide this struct as part
+ * of the ieee80211_tx_status. this struct should give information
+ * about the tx that is not part of the plcp rates table.
+ * @bitrate: rate used.
+ * @flags: Tx information (e.g. antenna selectd, guard interval, etc.)
+ * @retry_count: number of retries for this specific rate
+ */
+struct ieee80211_tx_rate_ctrl {
+       u32 bitrate;
+       u32 flags;
+       u8 retry_count;
+};
+
+/**
  * enum ieee80211_tx_status_flags - transmit status flags
  *
  * Status flags to indicate various transmit conditions.
@@ -343,7 +359,7 @@ enum ieee80211_tx_status_flags {
  * @control: a copy of the &struct ieee80211_tx_control passed to the driver
  *     in the tx() callback.
  * @flags: transmit status flags, defined above
- * @retry_count: number of retries
+ * @retry_count: total number of retries
  * @excessive_retries: set to 1 if the frame was retried many times
  *     but not acknowledged
  * @ampdu_ack_len: number of aggregated frames.
@@ -353,6 +369,8 @@ enum ieee80211_tx_status_flags {
  * @ack_signal: signal strength of the ACK frame
  * @queue_length: ?? REMOVE
  * @queue_number: ?? REMOVE
+ * @rate_ctrl_num: number of rate_ctrl allocated
+ * @rate_ctrl: info for rate at which the packet was actually transmitted
  */
 struct ieee80211_tx_status {
        struct ieee80211_tx_control control;
@@ -364,6 +382,8 @@ struct ieee80211_tx_status {
        int ack_signal;
        int queue_length;
        int queue_number;
+       size_t rate_ctrl_num;
+       struct ieee80211_tx_rate_ctrl rate_ctrl[0];
 };
 
Quoted 6 lines
Quoted 3 lines
include/net/mac80211.h +0 -1
 /**
--

1.5.2.2

Felix will this work for you?

Yes, looks good.

- Felix

maintainer
#3 lore raw
In reply to Tomas Winkler
Quoted 3 lines

+ * @flags: Tx information (e.g. antenna selectd, guard interval, etc.)

typo: selected, also where are the flags defined?

include/net/mac80211.h struct ieee80211_tx_status { · line 386

Quoted 8 lines

struct ieee80211_tx_status {

include/net/mac80211.h +2 -0
 	struct ieee80211_tx_control control;
@@ -364,6 +382,8 @@ struct ieee80211_tx_status {
 	int ack_signal;
 	int queue_length;
 	int queue_number;
+	size_t rate_ctrl_num;
+	struct ieee80211_tx_rate_ctrl rate_ctrl[0];

I think this is going to blow up if any driver sets rate_ctrl_num > 0 and uses the irq-safe tx status interface because we there copy the stuff into another buffer and you haven't changed that buffer to be resized accordingly.

johannes

#4 lore raw
Quoted 6 lines · Johannes Berg
Quoted 3 lines

+ * @flags: Tx information (e.g. antenna selectd, guard interval, etc.)

typo: selected, also where are the flags defined?

Quoted 13 lines

include/net/mac80211.h struct ieee80211_tx_status { · line 386

Quoted 8 lines
 struct ieee80211_tx_status {
      struct ieee80211_tx_control control;
include/net/mac80211.h +2 -0
@@ -364,6 +382,8 @@ struct ieee80211_tx_status {
      int ack_signal;
      int queue_length;
      int queue_number;
+     size_t rate_ctrl_num;
+     struct ieee80211_tx_rate_ctrl rate_ctrl[0];

I think this is going to blow up if any driver sets rate_ctrl_num > 0 and uses the irq-safe tx status interface because we there copy the stuff into another buffer and you haven't changed that buffer to be resized accordingly.

I have the same concern and wanted to set to some reasonable number but Felix thinks otherwise.
In general the driver should know how many rates it wants to report so in TX-response path it's allocates enough memory.
Since on the TX path tx_status resides on stack this has to be copied by the driver for TX-response path use.
Yes it's still , it is quite error prone :(, maybe some helper macros/functions can make it less vulnerable.

sizeof(struct ieee80211_tx_status) + rates_num * sizeof(struct ieee80211_tx_rate_ctrl)

void ieee80211_tx_status_irqsafe(struct ieee80211_hw *hw,

				 struct sk_buff *skb,
				 struct ieee80211_tx_status *status)
	

	skb->dev = local->mdev;
	saved = kmalloc(sizeof(struct ieee80211_tx_status), GFP_ATOMIC); ---
this has to be changed as well.
	if (unlikely(!saved)) {
Quoted 2 lines

johannes

maintainer
#5 lore raw
Quoted 2 lines

In general the driver should know how many rates it wants to report so in TX-response path it's allocates enough memory.

Yeah, obviously.

Quoted 2 lines

Since on the TX path tx_status resides on stack this has to be copied by the driver for TX-response path use.

I don't think I understand? This is tx status, what are you referring to by "tx response"?

Quoted 8 lines

void ieee80211_tx_status_irqsafe(struct ieee80211_hw *hw,

				 struct sk_buff *skb,
				 struct ieee80211_tx_status *status)
	

	skb->dev = local->mdev;
	saved = kmalloc(sizeof(struct ieee80211_tx_status), GFP_ATOMIC); ---

this has to be changed as well.

Yeah that's what I was thinking of. Maybe we should change the ieee80211_tx_status_irqsafe() interface and require the driver to allocate it rather than giving a pointer to a structure on the stack? That way we the driver has to do the calculation, but we don't have to copy things around? Doing it on the stack isn't easy, you'd have to define something like this:

struct status {
	struct ieee80211_tx_status stat;
	struct ieee80211_tx_rate_ctrl ratectrl[N];
} status __attribute__((packed));

so requiring the driver to allocate it when using the irqsafe interface would be better imho.

johannes

#6 lore raw
Quoted 5 lines
Quoted 2 lines

Since on the TX path tx_status resides on stack this has to be copied by the driver for TX-response path use.

I don't think I understand? This is tx status, what are you referring to by "tx response"?

Maybe I'm locked in concepts of iwlwifi but I believe also other drivers uses similar concept. I think USB has some URB complete handler etc. So iwlwifi get interrupt (asynchronous) when packet is transmitted (TX response flow) and driver retrieves the information of transmitting. This is NOT done in TX contexts so you have to store tx_status between TX flows and TX response flow so you can feed it back to ieee80211_tx_status_irqsafe

Quoted 18 lines
Quoted 9 lines

void ieee80211_tx_status_irqsafe(struct ieee80211_hw *hw,

                               struct sk_buff *skb,
                               struct ieee80211_tx_status *status)


      skb->dev = local->mdev;
      saved = kmalloc(sizeof(struct ieee80211_tx_status), GFP_ATOMIC); ---

this has to be changed as well.

Yeah that's what I was thinking of. Maybe we should change the ieee80211_tx_status_irqsafe() interface and require the driver to allocate it rather than giving a pointer to a structure on the stack? That way we the driver has to do the calculation, but we don't have to copy things around? Doing it on the stack isn't easy, you'd have to define something like this:

Quoted 8 lines
struct status {
        struct ieee80211_tx_status stat;
        struct ieee80211_tx_rate_ctrl ratectrl[N];
} status __attribute__((packed));

so requiring the driver to allocate it when using the irqsafe interface would be better imho.

Not sure which flow will free this pointer ? Iwlwifi keeps tx_status attached to TBD so it have to be sure it won't be overriden somewhere in the stack. But it might be a case that I'm not understand here something

Thanks
Tomas

Quoted 2 lines

johannes

maintainer
#7 lore raw
Quoted 8 lines

Maybe I'm locked in concepts of iwlwifi but I believe also other drivers uses similar concept. I think USB has some URB complete handler etc. So iwlwifi get interrupt (asynchronous) when packet is transmitted (TX response flow) and driver retrieves the information of transmitting. This is NOT done in TX contexts so you have to store tx_status between TX flows and TX response flow so you can feed it back to ieee80211_tx_status_irqsafe

Ah. But where's the difference between "TX response" and "TX status" then? Anyway, yes, you do have to store the info between ->tx() and tx_status().

Quoted 4 lines

Not sure which flow will free this pointer ? Iwlwifi keeps tx_status attached to TBD so it have to be sure it won't be overriden somewhere in the stack. But it might be a case that I'm not understand here something

I'd think that mac80211 would free it. I guess we should then have something like

static inline struct ieee80211_tx_status *
ieee80211_alloc_tx_status(size_t num_rate_ctrl)

{
	struct ieee80211_tx_status *s;
	s = kzalloc(... the expression you had earlier...);
	if (s)
		s->num_rate_control = num_rate_control;
	return s;
}

so that we can change that w/o changing all drivers. In fact, you'd probably already allocate that in the TX flow and keep it around until TX status.

Mind you, I'm just writing whatever comes to my head ;) Better ideas welcome, especially since this makes tx_status_irqsafe() and tx_status() behave completely differently. Incidentally, why does iwlwifi use both depending on in_interrupt(), or rather, why isn't the content statically known?

johannes

#8 lore raw
Quoted 11 lines · Johannes Berg
Quoted 8 lines

Maybe I'm locked in concepts of iwlwifi but I believe also other drivers uses similar concept. I think USB has some URB complete handler etc. So iwlwifi get interrupt (asynchronous) when packet is transmitted (TX response flow) and driver retrieves the information of transmitting. This is NOT done in TX contexts so you have to store tx_status between TX flows and TX response flow so you can feed it back to ieee80211_tx_status_irqsafe

Ah. But where's the difference between "TX response" and "TX status" then?

There are the same, just terminology juggling.

Anyway, yes, you do have to store the info between ->tx() and

Quoted 20 lines

tx_status().

Quoted 4 lines

Not sure which flow will free this pointer ? Iwlwifi keeps tx_status attached to TBD so it have to be sure it won't be overriden somewhere in the stack. But it might be a case that I'm not understand here something

I'd think that mac80211 would free it. I guess we should then have something like

static inline struct ieee80211_tx_status *
ieee80211_alloc_tx_status(size_t num_rate_ctrl)

{
        struct ieee80211_tx_status *s;
        s = kzalloc(... the expression you had earlier...);
        if (s)
                s->num_rate_control = num_rate_control;
        return s;
}
Quoted 3 lines

so that we can change that w/o changing all drivers. In fact, you'd probably already allocate that in the TX flow and keep it around until TX status.

It's allocated in <bus>_probe stage. The 'dynamic' allocation is done only by mac in ieee80211_tx_status_irqsafe.

Quoted 3 lines

Mind you, I'm just writing whatever comes to my head ;) Better ideas welcome, especially since this makes tx_status_irqsafe() and tx_status() behave completely differently.

I prefer to start to converge to something since iwlwifi driver is broken for week on the tip of everything.

Incidentally, why does iwlwifi use both

Quoted 2 lines

depending on in_interrupt(), or rather, why isn't the content statically known?

Actually in spite of the IF there I don't think it ever happens that iwlwifi uses tx_status() directly only irqsafe. Does any driver uses tx_status directly anyway?

Tomas

Tomas

Quoted 2 lines

johannes

maintainer
#9 lore raw
Quoted 7 lines
Quoted 4 lines

Mind you, I'm just writing whatever comes to my head ;) Better ideas welcome, especially since this makes tx_status_irqsafe() and tx_status() behave completely differently.

I prefer to start to converge to something since iwlwifi driver is broken for week on the tip of everything.

Yeah let's fix that and then do the larger redesign later.

Quoted 7 lines

Incidentally, why does iwlwifi use both

Quoted 2 lines

depending on in_interrupt(), or rather, why isn't the content statically known?

Actually in spite of the IF there I don't think it ever happens that iwlwifi uses tx_status() directly only irqsafe. Does any driver uses tx_status directly anyway?

ath5k does, but I have no idea why.

johannes

#10 lore raw
Quoted 6 lines
Quoted 3 lines

Actually in spite of the IF there I don't think it ever happens that iwlwifi uses tx_status() directly only irqsafe. Does any driver uses tx_status directly anyway?

ath5k does, but I have no idea why.

We have separate tasklets for tx/rx (so we can parse descriptrors, keep buffers/descriptors in sync etc), so we call ieee80211_tx_status from our tx tasklet.

Signature · 3 lines
GPG ID: 0xD21DB2DB
As you read this post global entropy rises. Have Fun ;-)
Nick