public inbox for netdev@vger.kernel.org
 help / color / mirror / Atom feed
From: Michal Schmidt <mschmidt@redhat.com>
To: "netdev@vger.kernel.org" <netdev@vger.kernel.org>
Cc: "Vlad Zolotarov" <vladz@broadcom.com>,
	"Dmitry Kravkov" <dmitry@broadcom.com>,
	"Eilon Greenstein" <eilong@broadcom.com>,
	"mirqus@gmail.com" <mirqus@gmail.com>
Subject: Re: [PATCH 1/3] bnx2x: remove TPA_ENABLE_FLAG
Date: Wed, 31 Aug 2011 18:55:07 +0200	[thread overview]
Message-ID: <20110831185507.50a1940c@alice> (raw)
In-Reply-To: <201108311816.30468.vladz@broadcom.com>

On Wed, 31 Aug 2011 18:16:30 +0300 Vlad Zolotarov wrote:
> This is bogus - what if bnx2x_reload_if_running(dev)
> (bnx2x_nic_load()) failes? The original dev->features would be
> lost... Pls., see my latest response on your previouse patch series
> thread.

OK, so the conclusion in the other thread was to restore the features
before returning. Here's an updated patch.


Subject: [PATCH] bnx2x: remove TPA_ENABLE_FLAG

TPA_ENABLE_FLAG is unnecessary, because it mirrors NETIF_F_LRO.

"Cheating" in bnx2x_set_features() was suggested by Michał Mirosław.

Signed-off-by: Michal Schmidt <mschmidt@redhat.com>
---
 drivers/net/ethernet/broadcom/bnx2x/bnx2x.h      |    2 +-
 drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c  |   31 ++++++++++++---------
 drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c |    7 +---
 3 files changed, 21 insertions(+), 19 deletions(-)

diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x.h b/drivers/net/ethernet/broadcom/bnx2x/bnx2x.h
index 735e491..5d5f323 100644
--- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x.h
+++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x.h
@@ -1174,7 +1174,7 @@ struct bnx2x {
 #define USING_MSIX_FLAG			(1 << 5)
 #define USING_MSI_FLAG			(1 << 6)
 #define DISABLE_MSI_FLAG		(1 << 7)
-#define TPA_ENABLE_FLAG			(1 << 8)
+
 #define NO_MCP_FLAG			(1 << 9)
 
 #define BP_NOMCP(bp)			(bp->flags & NO_MCP_FLAG)
diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c
index a08b101..399e71c 100644
--- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c
+++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_cmn.c
@@ -62,7 +62,7 @@ static inline void bnx2x_bz_fp(struct bnx2x *bp, int index)
 	 * set the tpa flag for each queue. The tpa flag determines the queue
 	 * minimal size so it must be set prior to queue memory allocation
 	 */
-	fp->disable_tpa = ((bp->flags & TPA_ENABLE_FLAG) == 0);
+	fp->disable_tpa = !(bp->dev->features & NETIF_F_LRO);
 
 #ifdef BCM_CNIC
 	/* We don't want TPA on an FCoE L2 ring */
@@ -3410,13 +3410,10 @@ u32 bnx2x_fix_features(struct net_device *dev, u32 features)
 int bnx2x_set_features(struct net_device *dev, u32 features)
 {
 	struct bnx2x *bp = netdev_priv(dev);
-	u32 flags = bp->flags;
 	bool bnx2x_reload = false;
 
-	if (features & NETIF_F_LRO)
-		flags |= TPA_ENABLE_FLAG;
-	else
-		flags &= ~TPA_ENABLE_FLAG;
+	if ((features ^ dev->features) & NETIF_F_LRO)
+		bnx2x_reload = true;
 
 	if (features & NETIF_F_LOOPBACK) {
 		if (bp->link_params.loopback_mode != LOOPBACK_BMAC) {
@@ -3430,14 +3427,22 @@ int bnx2x_set_features(struct net_device *dev, u32 features)
 		}
 	}
 
-	if (flags ^ bp->flags) {
-		bp->flags = flags;
-		bnx2x_reload = true;
-	}
-
 	if (bnx2x_reload) {
-		if (bp->recovery_state == BNX2X_RECOVERY_DONE)
-			return bnx2x_reload_if_running(dev);
+		if (bp->recovery_state == BNX2X_RECOVERY_DONE) {
+			/*
+			 * Cheat! Normally dev->features will be set after we
+			 * return, but that's too late. We need to know how to
+			 * configure the NIC when reloading it, so copy the
+			 * features beforehand. Restore them after reload to
+			 * avoid any possible confusion of the caller.
+			 */
+			int ret;
+			u32 orig_features = dev->features;
+			dev->features = features;
+			ret = bnx2x_reload_if_running(dev);
+			dev->features = orig_features;
+			return ret;
+		}
 		/* else: bnx2x_nic_load() will be called at end of recovery */
 	}
 
diff --git a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c
index e7b584b..fd32c04 100644
--- a/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c
+++ b/drivers/net/ethernet/broadcom/bnx2x/bnx2x_main.c
@@ -9757,13 +9757,10 @@ static int __devinit bnx2x_init_bp(struct bnx2x *bp)
 	bp->multi_mode = multi_mode;
 
 	/* Set TPA flags */
-	if (disable_tpa) {
-		bp->flags &= ~TPA_ENABLE_FLAG;
+	if (disable_tpa)
 		bp->dev->features &= ~NETIF_F_LRO;
-	} else {
-		bp->flags |= TPA_ENABLE_FLAG;
+	else
 		bp->dev->features |= NETIF_F_LRO;
-	}
 	bp->disable_tpa = disable_tpa;
 
 	if (CHIP_IS_E1(bp))
-- 
1.7.6

  parent reply	other threads:[~2011-08-31 17:45 UTC|newest]

Thread overview: 12+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2011-08-31 15:00 [PATCH 0/3 net-next] bnx2x: VLAN stripping toggle and a small cleanup Michal Schmidt
2011-08-31 15:00 ` [PATCH 1/3] bnx2x: remove TPA_ENABLE_FLAG Michal Schmidt
2011-08-31 15:16   ` Vlad Zolotarov
2011-08-31 15:58     ` Michal Schmidt
2011-08-31 16:13       ` Vladislav Zolotarov
2011-08-31 18:05         ` Michał Mirosław
2011-09-01  8:37           ` Vladislav Zolotarov
2011-09-01  9:37             ` Michał Mirosław
2011-09-01 10:52               ` Vladislav Zolotarov
2011-08-31 16:55     ` Michal Schmidt [this message]
2011-08-31 15:00 ` [PATCH 2/3] bnx2x: do not set NETIF_F_LRO in bnx2x_init_bp Michal Schmidt
2011-08-31 15:00 ` [PATCH 3/3] bnx2x: expose HW RX VLAN stripping toggle Michal Schmidt

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=20110831185507.50a1940c@alice \
    --to=mschmidt@redhat.com \
    --cc=dmitry@broadcom.com \
    --cc=eilong@broadcom.com \
    --cc=mirqus@gmail.com \
    --cc=netdev@vger.kernel.org \
    --cc=vladz@broadcom.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