Adding support for spectral scan for the ath9k_htc driver. Tested the patch on with 9271 chipset to generate spectral scans.

PATCH New Ashish Patro 3 participants 8 messages Apr 6, 2014 – Apr 14, 2014 Qualcomm Atheros ATH9K wireless driver

Share & tools

Ashish Patro

author

2014-04-06 23:23 UTC #0 lore raw

PATCH 1/1 Adding support for spectral scan for the ath9k_htc driver. Tested the patch on with 9271 chipset to generate spectral scans.

- Used the same mechanism to collect spectral scan data as used by the existing ath9k code.
- Minor refactoring done to the code to separate the ath9k and ath9k_htc related spectral scan functions into separate files.

This work was done at the WiNGS Lab at the University of Wisconsin Madison.

---
 drivers/net/wireless/ath/ath9k/Makefile         |    2 +-
 drivers/net/wireless/ath/ath9k/ar9002_phy.c     |   13 +-
 drivers/net/wireless/ath/ath9k/ath9k.h          |    1 +
 drivers/net/wireless/ath/ath9k/htc.h            |    7 +
 drivers/net/wireless/ath/ath9k/htc_drv_debug.c  |    3 +
 drivers/net/wireless/ath/ath9k/htc_drv_init.c   |    8 +
 drivers/net/wireless/ath/ath9k/htc_drv_main.c   |   81 ++++-
 drivers/net/wireless/ath/ath9k/htc_drv_txrx.c   |   14 +
 drivers/net/wireless/ath/ath9k/spectral.h       |   19 -
 drivers/net/wireless/ath/ath9k/spectral_ath9k.h |   39 ++
 drivers/net/wireless/ath/ath9k/spectral_htc.c   |  549 +++++++++++++++++++++++
 drivers/net/wireless/ath/ath9k/spectral_htc.h   |   40 ++
 12 files changed, 752 insertions(+), 24 deletions(-)
 create mode 100644 drivers/net/wireless/ath/ath9k/spectral_ath9k.h
 create mode 100644 drivers/net/wireless/ath/ath9k/spectral_htc.c
 create mode 100644 drivers/net/wireless/ath/ath9k/spectral_htc.h

drivers/net/wireless/ath/ath9k/Makefile +1 -1

index 8e1c7b0..56555a9 100644
@@ -64,6 +64,6 @@ ath9k_htc-y +=	htc_hst.o \
 		htc_drv_init.o \
 		htc_drv_gpio.o
 
-ath9k_htc-$(CONFIG_ATH9K_HTC_DEBUGFS) += htc_drv_debug.o
+ath9k_htc-$(CONFIG_ATH9K_HTC_DEBUGFS) += htc_drv_debug.o spectral_htc.o
 
 obj-$(CONFIG_ATH9K_HTC) += ath9k_htc.o

drivers/net/wireless/ath/ath9k/ar9002_phy.c +10 -3

index 9a2afa2..21340c9 100644
@@ -622,7 +622,7 @@ static void ar9002_hw_spectral_scan_config(struct ath_hw *ah,
 				    struct ath_spec_scan *param)
 {
 	u8 count;
-
+  
 	if (!param->enabled) {
 		REG_CLR_BIT(ah, AR_PHY_SPECTRAL_SCAN,
 			    AR_PHY_SPECTRAL_SCAN_ENABLE);

@@ -641,13 +641,20 @@ static void ar9002_hw_spectral_scan_config(struct ath_hw *ah,
 	/* on AR92xx, the highest bit of count will make the the chip send
 	 * spectral samples endlessly. Check if this really was intended,
 	 * and fix otherwise.
+   *
+   * For AR9271 chipsets, set count = 0 using the debugfs mechanism.
 	 */
 	count = param->count;
-	if (param->endless)
+	if (param->endless) {
 		count = 0x80;
-	else if (count & 0x80)
+
+    if (AR_SREV_9271(ah)) {
+      count = 0;
+    }
+	} else if (count & 0x80)
 		count = 0x7f;
 
+
 	REG_RMW_FIELD(ah, AR_PHY_SPECTRAL_SCAN,
 		      AR_PHY_SPECTRAL_SCAN_COUNT, count);
 	REG_RMW_FIELD(ah, AR_PHY_SPECTRAL_SCAN,

drivers/net/wireless/ath/ath9k/ath9k.h +1 -0

index 44d7449..8b59901 100644
@@ -28,6 +28,7 @@
 #include "mci.h"
 #include "dfs.h"
 #include "spectral.h"
+#include "spectral_ath9k.h"
 
 struct ath_node;

drivers/net/wireless/ath/ath9k/htc.h +7 -0

index dab1f0c..5fb778e 100644
@@ -31,6 +31,8 @@
 #include "htc_hst.h"
 #include "hif_usb.h"
 #include "wmi.h"
+#include "spectral.h"
+#include "spectral_htc.h"
 
 #define ATH_STA_SHORT_CALINTERVAL 1000    /* 1 second */
 #define ATH_AP_SHORT_CALINTERVAL  100     /* 100 ms */

@@ -529,6 +531,11 @@ struct ath9k_htc_priv {
 	struct ath9k_debug debug;
 #endif
 	struct mutex mutex;
+
+	/* relay(fs) channel for spectral scan */
+	struct rchan *rfs_chan_spec_scan;
+	enum spectral_mode spectral_mode;
+	struct ath_spec_scan spec_config;
 };
 
 static inline void ath_read_cachesize(struct ath_common *common, int *csz)

drivers/net/wireless/ath/ath9k/htc_drv_debug.c +3 -0

index fb071ee..e63bab3 100644
@@ -992,6 +992,9 @@ int ath9k_htc_init_debug(struct ath_hw *ah)
 					     priv->hw->wiphy->debugfsdir);
 	if (!priv->debug.debugfs_phy)
 		return -ENOMEM;
+  
+  /* Initialize the debugFS files for the spectral scans */
+	ath9k_htc_spectral_init_debug(priv);
 
 	debugfs_create_file("tgt_int_stats", S_IRUSR, priv->debug.debugfs_phy,
 			    priv, &fops_tgt_int_stats);

drivers/net/wireless/ath/ath9k/htc_drv_init.c +8 -0

index 8a3bd5f..71a80ba 100644
@@ -449,6 +449,14 @@ static void ath9k_init_misc(struct ath9k_htc_priv *priv)
 
 	common->last_rssi = ATH_RSSI_DUMMY_MARKER;
 	priv->ah->opmode = NL80211_IFTYPE_STATION;
+
+	priv->spec_config.enabled = 0;
+  /* Set false for AR9271. */
+	priv->spec_config.short_repeat = false;
+	priv->spec_config.count = 8;
+	priv->spec_config.endless = false;
+	priv->spec_config.period = 18;
+	priv->spec_config.fft_period = 0x02;
 }
 
 static int ath9k_init_priv(struct ath9k_htc_priv *priv,

drivers/net/wireless/ath/ath9k/htc_drv_main.c +80 -1

93 lines not on this page · raw message

drivers/net/wireless/ath/ath9k/htc_drv_txrx.c +14 -0

index e8149e3..d77a6b7 100644
@@ -1010,6 +1010,20 @@ static bool ath9k_rx_prepare(struct ath9k_htc_priv *priv,
 	 * separately to avoid doing two lookups for a rate for each frame.
 	 */
 	hdr = (struct ieee80211_hdr *)skb->data;
+
+	/*
+	 * Process PHY errors and return so that the packet
+	 * can be dropped.
+	 */
+	if (rx_stats.rs_status & ATH9K_RXERR_PHY) {
+    /* TODO: Not using DFS processing now. */
+		if (ath_htc_process_fft(priv, hdr, &rx_stats, rx_status->mactime)) {
+      /* TODO: Add code for collecting statistics about spectral scan */
+    }
+
+		goto rx_next;
+	}
+
 	if (!ath9k_cmn_rx_accept(common, hdr, rx_status, &rx_stats,
 			&decrypt_error, priv->rxfilter))
 		goto rx_next;

drivers/net/wireless/ath/ath9k/spectral.h +0 ~~-19~~

index ead6341..d321f7f 100644
@@ -190,23 +190,4 @@ struct fft_sample_ht20_40 {
 	u8 data[SPECTRAL_HT20_40_NUM_BINS];
 } __packed;
 
-void ath9k_spectral_init_debug(struct ath_softc *sc);
-void ath9k_spectral_deinit_debug(struct ath_softc *sc);
-
-void ath9k_spectral_scan_trigger(struct ieee80211_hw *hw);
-int ath9k_spectral_scan_config(struct ieee80211_hw *hw,
-			       enum spectral_mode spectral_mode);
-
-#ifdef CONFIG_ATH9K_DEBUGFS
-int ath_process_fft(struct ath_softc *sc, struct ieee80211_hdr *hdr,
-		    struct ath_rx_status *rs, u64 tsf);
-#else
-static inline int ath_process_fft(struct ath_softc *sc,
-				  struct ieee80211_hdr *hdr,
-				  struct ath_rx_status *rs, u64 tsf)
-{
-	return 0;
-}
-#endif /* CONFIG_ATH9K_DEBUGFS */
-
 #endif /* SPECTRAL_H */

drivers/net/wireless/ath/ath9k/spectral_ath9k.h new file +39 -0

new file mode 100644
index 0000000..08ae2c3
@@ -0,0 +1,39 @@
+/*
+ * Copyright (c) 2013 Qualcomm Atheros, Inc.
+ *
+ * Permission to use, copy, modify, and/or distribute this software for any
+ * purpose with or without fee is hereby granted, provided that the above
+ * copyright notice and this permission notice appear in all copies.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+ * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
+ * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+ * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+ * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
+ * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+ * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+ */
+
+#ifndef SPECTRAL_ATH9K_H
+#define SPECTRAL_ATH9K_H
+
+void ath9k_spectral_init_debug(struct ath_softc *sc);
+void ath9k_spectral_deinit_debug(struct ath_softc *sc);
+
+void ath9k_spectral_scan_trigger(struct ieee80211_hw *hw);
+int ath9k_spectral_scan_config(struct ieee80211_hw *hw,
+			       enum spectral_mode spectral_mode);
+
+#ifdef CONFIG_ATH9K_DEBUGFS
+int ath_process_fft(struct ath_softc *sc, struct ieee80211_hdr *hdr,
+		    struct ath_rx_status *rs, u64 tsf);
+#else
+static inline int ath_process_fft(struct ath_softc *sc,
+				  struct ieee80211_hdr *hdr,
+				  struct ath_rx_status *rs, u64 tsf)
+{
+	return 0;
+}
+#endif /* CONFIG_ATH9K_DEBUGFS */
+
+#endif /* SPECTRAL_ATH9K_H */

drivers/net/wireless/ath/ath9k/spectral_htc.c new file +549 -0

549 lines not on this page · raw message

drivers/net/wireless/ath/ath9k/spectral_htc.h new file +40 -0

new file mode 100644
index 0000000..546a8ac
@@ -0,0 +1,40 @@
+/*
+ * Copyright (c) 2013 Qualcomm Atheros, Inc.
+ *
+ * Permission to use, copy, modify, and/or distribute this software for any
+ * purpose with or without fee is hereby granted, provided that the above
+ * copyright notice and this permission notice appear in all copies.
+ *
+ * THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
+ * WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
+ * MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
+ * ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
+ * WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
+ * ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
+ * OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
+ */
+
+#ifndef SPECTRAL_HTC_H
+#define SPECTRAL_HTC_H
+
+/* Added a separate set of spectral scan functions for the ath9k_htc driver */
+void ath9k_htc_spectral_init_debug(struct ath9k_htc_priv *priv);
+void ath9k_htc_spectral_deinit_debug(struct ath9k_htc_priv *priv);
+
+void ath9k_htc_spectral_scan_trigger(struct ieee80211_hw *hw);
+int ath9k_htc_spectral_scan_config(struct ieee80211_hw *hw,
+			       enum spectral_mode spectral_mode);
+
+#ifdef CONFIG_ATH9K_HTC_DEBUGFS
+int ath_htc_process_fft(struct ath9k_htc_priv *priv, struct ieee80211_hdr *hdr,
+		    struct ath_rx_status *rs, u64 tsf);
+#else
+static inline int ath_htc_process_fft(struct ath9k_htc_priv *priv,
+				  struct ieee80211_hdr *hdr,
+				  struct ath_rx_status *rs, u64 tsf)
+{
+	return 0;
+}
+#endif /* CONFIG_ATH9K_HTC_DEBUGFS */
+
+#endif /* SPECTRAL_HTC_H */

Signature · 1 line

1.7.5.4

Oleksij Rempel

2014-04-07 13:47 UTC #1 lore raw

PATCH 1/1 Adding support for spectral scan for the ath9k_htc driver. Tested the patch on with 9271 chipset to generate spectral scans.

Am 07.04.2014 01:23, schrieb Ashish Patro:

Quoted 4 lines

- Used the same mechanism to collect spectral scan data as used by the existing ath9k code.
- Minor refactoring done to the code to separate the ath9k and ath9k_htc related spectral scan functions into separate files.
This work was done at the WiNGS Lab at the University of Wisconsin Madison.

Current patch produce lots of duplicated code with ath9k. It will increase work later. I guess you plan to fix? ;)

Signature · 10 lines

Regards,
Oleksij

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 278 bytes
Desc: OpenPGP digital signature
Url : http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20140407/1c53356a/attachment.pgp

Ashish Patro

2014-04-07 17:21 UTC #2 lore raw

PATCH 1/1 Adding support for spectral scan for the ath9k_htc driver. Tested the patch on with 9271 chipset to generate spectral scans.

Hi Oleksij,

I didn't want to poke too much into the ath9k part of the driver code, so I wrote it in the current way. All the duplicated code is currently present in spectral_htc.c. The debugFS related code needs to be mostly duplicated to provide the same interface for ath9k and ath9k_htc drivers.

The common code can be later moved into ath_common. The main issue is that the current ath9k code is using references to ath_softc, so the common code needs to be written to be independent of ath_softc and ath9k_htc_priv data structures.

Thanks
Ashish

Quoted 17 lines · Oleksij Rempel

Am 07.04.2014 01:23, schrieb Ashish Patro:
Quoted 1 line
- Used the same mechanism to collect spectral scan data as used by the
existing ath9k code.
Quoted 1 line
- Minor refactoring done to the code to separate the ath9k and ath9k_htc
related spectral scan functions into separate files.
Quoted 2 lines
This work was done at the WiNGS Lab at the University of Wisconsin
Madison.
Current patch produce lots of duplicated code with ath9k. It will increase work later. I guess you plan to fix? ;)
--
Regards,
Oleksij

Quoted 6 lines

_______________________________________________
ath9k-devel mailing list
ath9k-devel at lists.ath9k.org
https://lists.ath9k.org/mailman/listinfo/ath9k-devel

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20140407/225fb056/attachment.htm

Oleksij Rempel

2014-04-07 18:10 UTC #3 lore raw

PATCH 1/1 Adding support for spectral scan for the ath9k_htc driver. Tested the patch on with 9271 chipset to generate spectral scans.

Am 07.04.2014 19:21, schrieb Ashish Patro:

Quoted 11 lines

Hi Oleksij,
I didn't want to poke too much into the ath9k part of the driver code, so I wrote it in the current way. All the duplicated code is currently present in spectral_htc.c. The debugFS related code needs to be mostly duplicated to provide the same interface for ath9k and ath9k_htc drivers.
The common code can be later moved into ath_common. The main issue is that the current ath9k code is using references to ath_softc, so the common code needs to be written to be independent of ath_softc and ath9k_htc_priv data structures.

Correct. spectral.c has no too deep ath_softc dependencies, it can be solved.
I don't like the idea of having patch which i already know, that i will need rework it. Ath9k-htc has not enough developers, the goal is to reduce work, not increase it.

Quoted 26 lines

On Mon, Apr 7, 2014 at 8:47 AM, Oleksij Rempel <[email protected] <mailto:[email protected]>> wrote:
Am 07.04.2014 01:23, schrieb Ashish Patro:
> - Used the same mechanism to collect spectral scan data as used by the existing ath9k code.
> - Minor refactoring done to the code to separate the ath9k and ath9k_htc related spectral scan functions into separate files. >
> This work was done at the WiNGS Lab at the University of Wisconsin Madison.
Current patch produce lots of duplicated code with ath9k. It will increase work later. I guess you plan to fix? ;)
--
Regards,
Oleksij
_______________________________________________
ath9k-devel mailing list
ath9k-devel at lists.ath9k.org <mailto:[email protected]>
https://lists.ath9k.org/mailman/listinfo/ath9k-devel

Signature · 10 lines

Regards,
Oleksij

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 278 bytes
Desc: OpenPGP digital signature
Url : http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20140407/03c50324/attachment.pgp

Ashish Patro

2014-04-08 07:51 UTC #4 lore raw

PATCH 1/1 Adding support for spectral scan for the ath9k_htc driver. Tested the patch on with 9271 chipset to generate spectral scans.

Hi Oleksij,

I can take a pass at the code to reduce the duplicate code between ath9k and ath9k_htc. I?ll need to make changes to the current spectral.c code and add code into ath_common. Will it be ok for me to do this code refactoring? I?ve so far avoided doing significant refactoring of the existing code in the ath9k driver.

Thanks
Ashish

Quoted 51 lines · Oleksij Rempel

Am 07.04.2014 19:21, schrieb Ashish Patro:
Quoted 11 lines
Hi Oleksij,
I didn't want to poke too much into the ath9k part of the driver code, so I wrote it in the current way. All the duplicated code is currently present in spectral_htc.c. The debugFS related code needs to be mostly duplicated to provide the same interface for ath9k and ath9k_htc drivers.
The common code can be later moved into ath_common. The main issue is that the current ath9k code is using references to ath_softc, so the common code needs to be written to be independent of ath_softc and ath9k_htc_priv data structures.
Correct. spectral.c has no too deep ath_softc dependencies, it can be solved.
I don't like the idea of having patch which i already know, that i will need rework it. Ath9k-htc has not enough developers, the goal is to reduce work, not increase it.
Quoted 26 lines
On Mon, Apr 7, 2014 at 8:47 AM, Oleksij Rempel <[email protected] <mailto:[email protected]>> wrote:
   Am 07.04.2014 01:23, schrieb Ashish Patro:
Quoted 1 line
- Used the same mechanism to collect spectral scan data as used by
the existing ath9k code.
Quoted 1 line
- Minor refactoring done to the code to separate the ath9k and
ath9k_htc related spectral scan functions into separate files.
Quoted 2 lines
This work was done at the WiNGS Lab at the University of Wisconsin
Madison.
Current patch produce lots of duplicated code with ath9k. It will increase work later. I guess you plan to fix? ;)
--
Regards,
Oleksij
_______________________________________________
ath9k-devel mailing list
ath9k-devel at lists.ath9k.org <mailto:[email protected]>
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Signature · 2 lines
Regards,
Oleksij

Oleksij Rempel

2014-04-08 08:16 UTC #5 lore raw

PATCH 1/1 Adding support for spectral scan for the ath9k_htc driver. Tested the patch on with 9271 chipset to generate spectral scans.

Am 08.04.2014 09:51, schrieb Ashish Patro:

Quoted 3 lines

Hi Oleksij,
I can take a pass at the code to reduce the duplicate code between ath9k and ath9k_htc. I?ll need to make changes to the current spectral.c code and add code into ath_common. Will it be ok for me to do this code refactoring? I?ve so far avoided doing significant refactoring of the existing code in the ath9k driver.

It will be great if you 'll do it. If patches are ok, they will be accepted. Please separate all you stages to different patches. It will make bisection easier. End result should looks some way like this:
PATCH 1 - move spectralscan related data from ath_softc to some common struct and make all related changes in ath9k driver.
PATCH 2 - move all common spectral related functions to common-spectral.c PATCH 3 - make use of common-spectral.c in ath9k_htc.

After each patch source should be compilable and driver should work.

Quoted 57 lines

Thanks
Ashish
Quoted 51 lines · Oleksij Rempel
Am 07.04.2014 19:21, schrieb Ashish Patro:
Quoted 11 lines
Hi Oleksij,
I didn't want to poke too much into the ath9k part of the driver code, so I wrote it in the current way. All the duplicated code is currently present in spectral_htc.c. The debugFS related code needs to be mostly duplicated to provide the same interface for ath9k and ath9k_htc drivers.
The common code can be later moved into ath_common. The main issue is that the current ath9k code is using references to ath_softc, so the common code needs to be written to be independent of ath_softc and ath9k_htc_priv data structures.
Correct. spectral.c has no too deep ath_softc dependencies, it can be solved.
I don't like the idea of having patch which i already know, that i will need rework it. Ath9k-htc has not enough developers, the goal is to reduce work, not increase it.
Quoted 26 lines
On Mon, Apr 7, 2014 at 8:47 AM, Oleksij Rempel <[email protected] <mailto:[email protected]>> wrote:
   Am 07.04.2014 01:23, schrieb Ashish Patro:
Quoted 1 line
- Used the same mechanism to collect spectral scan data as used by
the existing ath9k code.
Quoted 1 line
- Minor refactoring done to the code to separate the ath9k and
ath9k_htc related spectral scan functions into separate files.
Quoted 2 lines
This work was done at the WiNGS Lab at the University of Wisconsin
Madison.
Current patch produce lots of duplicated code with ath9k. It will increase work later. I guess you plan to fix? ;)
--
Regards,
Oleksij
_______________________________________________
ath9k-devel mailing list
ath9k-devel at lists.ath9k.org <mailto:[email protected]>
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Signature · 2 lines
Regards,
Oleksij

Signature · 10 lines

Regards,
Oleksij

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 278 bytes
Desc: OpenPGP digital signature
Url : http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20140408/4c5fcdc4/attachment-0001.pgp

Ashish Patro

2014-04-14 18:32 UTC #6 lore raw

PATCH 1/1 Adding support for spectral scan for the ath9k_htc driver. Tested the patch on with 9271 chipset to generate spectral scans.

Hi Oleksij,

I've resubmitted three new patches to refactor the code and spectral scan support to ath9k_htc as discussed earlier.

Thanks
Ashish

Quoted 81 lines · Oleksij Rempel

Am 08.04.2014 09:51, schrieb Ashish Patro:
Quoted 3 lines
Hi Oleksij,
I can take a pass at the code to reduce the duplicate code between ath9k
and ath9k_htc. I?ll need to make changes to the current spectral.c code and add code into ath_common. Will it be ok for me to do this code refactoring? I?ve so far avoided doing significant refactoring of the existing code in the ath9k driver.
It will be great if you 'll do it. If patches are ok, they will be accepted.
Please separate all you stages to different patches. It will make bisection easier. End result should looks some way like this:
PATCH 1 - move spectralscan related data from ath_softc to some common struct and make all related changes in ath9k driver.
PATCH 2 - move all common spectral related functions to common-spectral.c PATCH 3 - make use of common-spectral.c in ath9k_htc.
After each patch source should be compilable and driver should work.
Quoted 4 lines
Thanks
Ashish
On Apr 7, 2014, at 1:10 PM, Oleksij Rempel <[email protected]>
Quoted 8 lines · wrote:
Quoted 8 lines
Am 07.04.2014 19:21, schrieb Ashish Patro:
Quoted 6 lines
Hi Oleksij,
I didn't want to poke too much into the ath9k part of the driver code, so I wrote it in the current way. All the duplicated code is currently present in spectral_htc.c. The debugFS related code needs to be mostly duplicated to provide the same interface for ath9k and ath9k_htc
drivers.
Quoted 40 lines
Quoted 40 lines
Quoted 5 lines
The common code can be later moved into ath_common. The main issue is that the current ath9k code is using references to ath_softc, so the common code needs to be written to be independent of ath_softc and ath9k_htc_priv data structures.
Correct. spectral.c has no too deep ath_softc dependencies, it can be solved.
I don't like the idea of having patch which i already know, that i will need rework it. Ath9k-htc has not enough developers, the goal is to reduce work, not increase it.
Quoted 26 lines
On Mon, Apr 7, 2014 at 8:47 AM, Oleksij Rempel <[email protected] <mailto:[email protected]>> wrote:
   Am 07.04.2014 01:23, schrieb Ashish Patro:
Quoted 1 line
- Used the same mechanism to collect spectral scan data as used by
the existing ath9k code.
Quoted 1 line
- Minor refactoring done to the code to separate the ath9k and
ath9k_htc related spectral scan functions into separate files.
Quoted 2 lines
This work was done at the WiNGS Lab at the University of Wisconsin
Madison.
Current patch produce lots of duplicated code with ath9k. It will increase work later. I guess you plan to fix? ;)
--
Regards,
Oleksij
_______________________________________________
ath9k-devel mailing list
ath9k-devel at lists.ath9k.org <mailto:[email protected]>
https://lists.ath9k.org/mailman/listinfo/ath9k-devel
Quoted 5 lines
Quoted 4 lines
--
Regards,
Oleksij

Quoted 5 lines

--
Regards,
Oleksij

-------------- next part --------------
An HTML attachment was scrubbed...
URL: http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20140414/ff41643a/attachment.htm

Oleksij Rempel

2014-04-14 20:23 UTC #7 lore raw

PATCH 1/1 Adding support for spectral scan for the ath9k_htc driver. Tested the patch on with 9271 chipset to generate spectral scans.

Am 14.04.2014 20:32, schrieb Ashish Patro:

Quoted 8 lines

Hi Oleksij,
I've resubmitted three new patches to refactor the code and spectral scan support to ath9k_htc as discussed earlier.
Thanks
Ashish

Thank you,
i'll write some comments.

Quoted 70 lines

On Tue, Apr 8, 2014 at 3:16 AM, Oleksij Rempel <[email protected] <mailto:[email protected]>> wrote:
Am 08.04.2014 09:51, schrieb Ashish Patro:
> Hi Oleksij,
>
> I can take a pass at the code to reduce the duplicate code between ath9k and ath9k_htc. I?ll need to make changes to the current spectral.c code and add code into ath_common. Will it be ok for me to do this code refactoring? I?ve so far avoided doing significant refactoring of the existing code in the ath9k driver.
It will be great if you 'll do it. If patches are ok, they will be accepted.
Please separate all you stages to different patches. It will make bisection easier. End result should looks some way like this:
PATCH 1 - move spectralscan related data from ath_softc to some common struct and make all related changes in ath9k driver.
PATCH 2 - move all common spectral related functions to common-spectral.c
PATCH 3 - make use of common-spectral.c in ath9k_htc.
After each patch source should be compilable and driver should work.
> Thanks
> Ashish
>
> On Apr 7, 2014, at 1:10 PM, Oleksij Rempel <[email protected] <mailto:[email protected]>> wrote:
>
>> Am 07.04.2014 19:21, schrieb Ashish Patro:
>>> Hi Oleksij,
>>>
>>> I didn't want to poke too much into the ath9k part of the driver code,
>>> so I wrote it in the current way. All the duplicated code is currently
>>> present in spectral_htc.c. The debugFS related code needs to be mostly
>>> duplicated to provide the same interface for ath9k and ath9k_htc drivers.
>>>
>>> The common code can be later moved into ath_common. The main issue is
>>> that the current ath9k code is using references to ath_softc, so the >>> common code needs to be written to be independent of ath_softc and >>> ath9k_htc_priv data structures.
>>
>> Correct. spectral.c has no too deep ath_softc dependencies, it can be >> solved.
>> I don't like the idea of having patch which i already know, that i will
>> need rework it. Ath9k-htc has not enough developers, the goal is to >> reduce work, not increase it.
>>
>>> On Mon, Apr 7, 2014 at 8:47 AM, Oleksij Rempel
<linux at rempel-privat.de <mailto:[email protected]> >>> <mailto:linux at rempel-privat.de <mailto:[email protected]>>> wrote:
>>>
>>> Am 07.04.2014 01:23, schrieb Ashish Patro:
>>>> - Used the same mechanism to collect spectral scan data as used by
>>> the existing ath9k code.
>>>> - Minor refactoring done to the code to separate the ath9k and
>>> ath9k_htc related spectral scan functions into separate files.
>>>>
>>>> This work was done at the WiNGS Lab at the University of Wisconsin
>>> Madison.
>>>

Signature · 10 lines

Regards,
Oleksij

-------------- next part --------------
A non-text attachment was scrubbed...
Name: signature.asc
Type: application/pgp-signature
Size: 278 bytes
Desc: OpenPGP digital signature
Url : http://lists.ath9k.org/pipermail/ath9k-devel/attachments/20140414/e3766eeb/attachment.pgp