From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from smtpout-03.galae.net (smtpout-03.galae.net [185.246.85.4]) (using TLSv1.2 with cipher ECDHE-RSA-AES256-GCM-SHA384 (256/256 bits)) (No client certificate requested) by smtp.subspace.kernel.org (Postfix) with ESMTPS id 5C3FB47A87E; Wed, 30 Sep 2026 22:54:03 +0000 (UTC) Authentication-Results: smtp.subspace.kernel.org; arc=none smtp.client-ip=185.246.85.4 ARC-Seal:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1790808852; cv=none; b=jnwRShd/2z9W0bEhqitU7JiXGpp3urOU8aQgFTZPf6sTaCJh01kqkU3CLfrdpia1pvgsKE34xHnodKhBRXv8m7j8V2KqfJYvopigdBteXvSdTpzpn1WpGkys8TWFoKO4HI/RwKuGAKvvCO6LgL0oRUHnTtRdGdywqf68W6TzUac= ARC-Message-Signature:i=1; a=rsa-sha256; d=subspace.kernel.org; s=arc-20240116; t=1790808852; c=relaxed/simple; bh=etu7DWmWHK8H43UqYhPtJ6biB6o6AmTRbJjC/S+zpLc=; h=From:Subject:Date:Message-Id:MIME-Version:Content-Type:To:Cc; b=d9RfPKaTFVI5B5lbMyz+W4kAJW9FpOYzU/Hg3MX//FvFc/vlyZAOpOn3vEpIQwNYN7B3wc6g0yk5sLRYWClaFqUBxaChxr7jjp9n9fhqYUrMYlKtsdkU7Jkrm5LquwtpC6vILd6h0qylO1/XldROypQzk1fD+zN2MrtMrZObCG0= ARC-Authentication-Results:i=1; smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=bootlin.com; spf=pass smtp.mailfrom=bootlin.com; dkim=pass (2048-bit key) header.d=bootlin.com header.i=@bootlin.com header.b=ZUv2t/VA; arc=none smtp.client-ip=185.246.85.4 Authentication-Results: smtp.subspace.kernel.org; dmarc=pass (p=reject dis=none) header.from=bootlin.com Authentication-Results: smtp.subspace.kernel.org; spf=pass smtp.mailfrom=bootlin.com Authentication-Results: smtp.subspace.kernel.org; dkim=pass (2048-bit key) header.d=bootlin.com header.i=@bootlin.com header.b="ZUv2t/VA" Received: from smtpout-01.galae.net (smtpout-01.galae.net [212.83.139.233]) by smtpout-03.galae.net (Postfix) with ESMTPS id 8E9234E410DE; Wed, 30 Sep 2026 22:54:00 +0000 (UTC) Received: from mail.galae.net (mail.galae.net [212.83.136.155]) by smtpout-01.galae.net (Postfix) with ESMTPS id 343F960749; Wed, 30 Sep 2026 22:54:00 +0000 (UTC) Received: from [127.0.0.1] (localhost [127.0.0.1]) by localhost (Mailerdaemon) with ESMTPSA id 2961A1032A3D5; Thu, 1 Oct 2026 00:53:47 +0200 (CEST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=bootlin.com; s=dkim; t=1790808835; h=from:subject:date:message-id:to:cc:mime-version:content-type: content-transfer-encoding; bh=iXWBJ/cy+vut2uwj1eb2gv6tpL4hILbqmQIPqICbKBo=; b=ZUv2t/VAp/awhjlJZZGRlBAm1J1+bgUnnb3IIq7QZj1fdBJ0b62jYWoKHGyl5DdQVSZBP5 56HcBxTANwHWPEpSoF+ktq3RF3fSnwJVkvq00E2Otl0kBe61gvFcZSoTzDTAUGYDRlkGFk pBvRHP1Xy0tC4A6dkm0fXleMw14tdxYexemmNpF+v1r1MbtCuu0drFZYuMFLWIKdERDZQz flX5TiV44jv1aG4y6lWyHdqaQV3BMCkYvyWzcPlg9AP0SKxaX+ahyCrOL8KhH2oB1CzIF5 6mw0L3eNvFHCEd8gcDT1g7ZLRN2JcmfdqL4Ovg5H+gnxYEfA31Tbp8Fva0qX5w== From: =?utf-8?q?Th=C3=A9o_Lebrun?= Subject: [PATCH net-next v10 0/8] net: macb: implement context swapping Date: Thu, 01 Oct 2026 00:53:44 +0200 Message-Id: <20261001-macb-context-v10-0-beb541bbb7df@bootlin.com> Precedence: bulk X-Mailing-List: linux-kernel@vger.kernel.org List-Id: List-Subscribe: List-Unsubscribe: MIME-Version: 1.0 Content-Type: text/plain; charset="utf-8" Content-Transfer-Encoding: 8bit X-B4-Tracking: v=1; b=H4sIAAAAAAAC/23RS07DMBAG4KtUXmM040dsd8U9EAs/qRFNUBKio qp3xwmLBrtLz8w3/0i+kimOOU7keLiSMS55ykNfHghPB+JPtn+PNIdSIAxYBwKQnq131A/9HC8 zdQG8TQwEikAK+Rpjypdt3yvp40z7MkXe/jrTt/uIfl63rbOnPM3D+LNFL7iJxyELUqBGGvDSO sUUe3HDMH/m/tkP5235wnYcoeKscJu4SQqDANe1nN+5atJ54VBaOkiHEh6kix1HVXGx8lKJKiH 44Foud5yJisvCZWccasmcV6nl3Y7z+viucGGCtCi4CUK3XN25Bl5xtfKggrAJhBam5XrPZcV14 c5zYCkFRCVabnYcWcVN4SoEl2T5OhnhP7/dbr9QhqdYvwIAAA== X-Change-ID: 20260401-macb-context-bd0caf20414d To: =?utf-8?q?Th=C3=A9o_Lebrun?= , Conor Dooley , Andrew Lunn , "David S. Miller" , Eric Dumazet , Jakub Kicinski , Paolo Abeni , Richard Cochran , Russell King Cc: netdev@vger.kernel.org, linux-kernel@vger.kernel.org, Paolo Valerio , Nicolai Buchwitz , Vladimir Kondratiev , Gregory CLEMENT , =?utf-8?q?Beno=C3=AEt_Monin?= , Tawfik Bayouk , Thomas Petazzoni X-Mailer: b4 0.15.2 X-Last-TLS-Session-Version: TLSv1.3 MACB has a pretty primitive approach to buffer management. They are all stored in `struct macb *bp`. On operations that require buffer realloc (set_ringparam & change_mtu at the moment), the only option is to close the interface, change our global state and re-open the interface. Two issues: - It doesn't fly on memory pressured systems; we free our precious buffers and don't manage to reallocate fully, meaning our machine just lost its network access. - Anecdotally, it is pretty slow because it implies a full PHY reinit. Instead, we shall: - allocate a new context (including buffers) first - if it fails, early return without any impact to the interface - stop interface - update global state (bp, netdev, etc) - pass newly allocated buffer pointers to the hardware - start interface - free old context This is what we implement here. Both .set_ringparam() and .ndo_change_mtu() are covered by this series. In the future, at least .set_channels() [0], XDP [1] and XSK [2] would benefit. --- We now rebased onto the quiesce work that's made it to net-next. Also, I shortened the series by dropping one useless patch (took me a while to realise!) and one that has been posted separately [4]. A Sashiko review also lead to another separate patch sent [5]. I wonder if we can get those changes in before the next merge window? The idea being that we can restart iterating on XDP [1] & XSK [2] during the next kernel cycle, bringing those long awaited features into MACB. Thanks, Have a nice day, Théo [0]: https://lore.kernel.org/netdev/20260317-macb-set-channels-v4-0-1bd4f4ffcfca@bootlin.com/ [1]: https://lore.kernel.org/netdev/20260323221047.2749577-1-pvalerio@redhat.com/ [2]: https://lore.kernel.org/netdev/20260304-macb-xsk-v1-0-ba2ebe2bdaa3@bootlin.com/ [3]: https://lore.kernel.org/all/DJXGIM9EGPT8.4UHNYP2Y0GOP@bootlin.com/ [4]: https://lore.kernel.org/netdev/20260930-macb-irq-v1-1-8994a4c8f3bb@bootlin.com/ [5]: https://lore.kernel.org/netdev/20260918-macb-ncr-rmw-v1-1-30c3494f93cd@bootlin.com/ --- Changes in v10: - Rebase onto latest net-next/main (47a144672573), dropping first 7 patches applied by Jakub. It also means we reuse the quiesce work introduced by 4fcaed2cb1d8 ("net: macb: quiesce IRQs and drain BH on interface close"). - Drop wrong patch "read ISR inside bp->lock critical section". - Patch "move printk() calls out of bp->lock critical section" was sent separately as it is standalone. https://lore.kernel.org/netdev/20260930-macb-irq-v1-1-8994a4c8f3bb@bootlin.com/ - In commit "change caps helpers signatures", in macb_rx_pending(), convert missed macb_is_gem() call from bp to bp->info. Restore bisectability. - Writes to bp->ctx are now made with WRITE_ONCE() + wmb(). - macb_tx_restart() now early exits if context swap is ongoing, to avoid races. For that, we grow its existing bp->lock critical section. - Link to v9: https://patch.msgid.link/20260812-macb-context-v9-0-7ddbf5f715e0@bootlin.com Changes in v9: - Drop unused `netdev` stack variable in macb_interrupt_misc(). - Rebase onto latest net-next/main (ac155a26750a), nothing to report. - Link to v8: https://patch.msgid.link/20260805-macb-context-v8-0-bc302ffd1174@bootlin.com Changes in v8: - Avoid growing existing race condition window inbetween macb_tx_error_task() and macb_interrupt() by moving printks. - On swap end, don't start tx queues if link-down. - On swap end, wake tx queues instead of only starting them. - Rebase onto latest net-next/main (a23b36233d41), nothing to report. - Link to v7: https://patch.msgid.link/20260803-macb-context-v7-0-4d7d4af04849@bootlin.com Changes in v7: - queue->tx_{head,tail} got renamed: fixup queue->tx_ptr_lock field comment and netdev_dbg() call in macb_start_xmit(). - Ensure bp->configured_{rx,tx}_ring_size have correct values in the EMAC case so that get_ringparam doesn't lie. - Refuse set_ringparam on EMAC, which is invalid (sizes are hardcoded). - Fix whitespace issue in "change function signatures to take contexts". - Fix rebase mistake: macb_tx_unmap() signature change moved from patch "change function signatures to take contexts" to "change caps helpers signatures". - Patch "move printk() calls out of bp->lock critical section": - Undrop netdev_vdbg() in macb_tx_error_task(); let's ignore the race for debug printk calls. - Move netdev_err() from macb_interrupt_misc() into macb_interrupt() so it can live outside the bp->lock critical section. - Add details to commit message. - Patch "use context swapping in .set_ringparam()": - Re-schedule LPI after swap with delay=1s. Pick 1s so that if link-up -> swap instantly, we still respect the IEEE spec. - Add bp->link_up set under bp->mac_cfg_lock. netif_carrier_ok() bit is unsafe because not set inside a critical section we can enter. - Drop mac_cfg_lock from macb_hresp_error_task(); made useless by above. - Move macb_halt_tx() out of bp->lock to avoid 14ms latency spike possibility. Safe to do so. - Patch "use context swapping in .ndo_change_mtu()": - Clarify commit message about EMAC. - Rebase onto latest net-next/main (69963a0678a3), nothing to report. - Link to v6: https://patch.msgid.link/20260731-macb-context-v6-0-49d5a1439d48@bootlin.com Changes in v6: - Non Sashiko based: - Take 3x Reviewed-By: Nicolai. - Silence checkpatch warnings in [02/15] about printk(..., __func__). - Rebase to latest net-next/main (a5c6ae8de11e), nothing to report. - Fix netconsole/netpoll deadlock coming from macb_tx_error_task() calling printk() inside the bp->lock critical section. - Move at91ether synchronize_irq() on close to protect more kfree(). - Clarify commit message for [11/15]: functions operating on the interface have no risk of using the wrong interface. - On swap end, only re-enable the interface if !netif_carrier_ok(). - macb_hresp_error_task() grabs bp->mac_cfg_lock as it toggles the carrier state, to avoid race with context swap. - On swap end, move netif_tx_start_all_queues() into bp->mac_cfg_lock. - Link to v5: https://patch.msgid.link/20260724-macb-context-v5-0-569b1852bc7f@bootlin.com Changes in v5: - Fix build on [PATCH 02/15]; `bp->dev` got renamed to `bp->netdev`. - Fix one word in commit message of [PATCH 07/15]. - Move bp->ctx->info assignment in at91ether_open() from [PATCH 12/15] to [PATCH 09/15]. - In [PATCH 14/15], reset DQL (dynamic queue limits) after tx disable, not before. - Take 7 Reviewed-by: Nicolai (weirdly not all detected by b4). - Rebase on latest net-next/main (89d8006259b8), nothing to report. - Link to v4: https://patch.msgid.link/20260717-macb-context-v4-0-0acbe7f10cdb@bootlin.com Changes in v4: - Disable tx_error_task then disable NAPI because error task does a napi_disable() which deadlocks if NAPI is already disabled. - Disable NAPI then disable tx_lpi_work, because NAPI might re-arm the latter. - Last iteration did mask-irqs then disable-and-wait-for-bh then disable-hw. This is flawed because BH rearm IRQs once done. We don't have a good way to signal to them they shouldn't do so (we don't want to lock from NAPI context). So instead we add a bp->ctx_swap flag, shielded by bp->lock, to indicate to our IRQ handler to ignore IRQs and self-disarm. - at91ether_close(): synchronize_irq() before freeing context. - macb_interrupt(): drop double ISR read (outside & inside bp->lock). - Rebase on latest net-next/main (f6f3b36c15ed). macb_free() since commit 27f575836cfe ("net: macb: drop in-flight Tx SKBs on close") needs access to queue stats => add them to macb_info. - Take trailers. - I did NOT init ctx->rx_ring_size/tx_ring_size/rx_buffer_size from at91ether_open(), as recommended by the netdev LLM. They are not the first fields in MACB that are present in both instances and that are uninitialised in one case. AT91 is almost a different driver. [3] - Link to v3: https://patch.msgid.link/20260701-macb-context-v3-0-00268d5b1502@bootlin.com Changes in v3: - Use `const struct macb_info *info` instead of bare `u32 caps` as helper arguments, for type safety. - macb_interrupt(): the pre-lock readl(ISR) to detect spurious interrupts is only done if CLEAR_ON_WRITE. - Don't forget allocating context in at91ether_open(). - swap: - Refuse swap for EMAC HW; it would crash because codepaths are so different. - Grab new bp->mac_cfg_lock to serialise with phylink MAC callbacks. We cannot rely on phydev->lock because it isn't present in the SFP or fixed-link cases. We also want to avoid phylink_stop() which triggers a slow PHY retrain. - swap start: - We used to do disable-irqs-and-hw then drain-all-bh-features, but then HW might be raced against. Instead we disable-irqs then drain-all-bh then disable-hw which means at disable-hw step no BH context can be active. - Use macb_halt_tx() helper to properly stop HW. - Disable BH features before netif_tx_disable() call to avoid queue wakeup races. - Use macb_queue_isr_clear() helper instead of manual if-then-writel. - swap end: - Grab bp->lock for the hardware reinit sequence composed of DMACFG and NCR writes. - Drop now useless EMAC check (we refuse EMAC HW before swapping). - nits: - New patch to rename macb_{alloc,free}_consistent() which don't only allocate consistent buffers since a long time ago. - Fix the start_xmit verbose netdev_vdbg() format string from %hu to %u because the queue index type changed. - Strong commit reword from "unify `struct macb *` naming convention" to "unify variable naming convention in at91ether functions" which was underselling the changes. - Rebase upon latest net-next/main (1c664ec4b9ea). - Link to v2: https://patch.msgid.link/20260410-macb-context-v2-0-af39f71d40b6@bootlin.com Changes in v2: - Patch "add subset of `struct macb` to `struct macb_context`" was messed up. It contained much more than what the name implied. Split into three commits (I caused trouble by rebase reordering). - Fix tieoff; V1 allocated it without initialisation. - Fix NULL pointer dereference on context in mab_get_regs() and macb_get_ringparam() when interface is offline. - Patch "unify device pointer naming convention": - Fix build issue when CONFIG_NETCONSOLE=y. - Rename `struct net_device *dev` to `netdev` in macb.h. - Rename `struct phy_device *phy` to `phydev` in macb_main.c. - On swap, call netdev_tx_reset_queue() to reset all DQL counters. - At end of swap, add missing kfree(old_ctx). - During HW disabling in swap, grab bp->lock to protect against IRQ handler. - On swap, cancel the three BH features MACB has: bp->hresp_err_bh_work, bp->tx_lpi_work and queue->tx_error_task. - On swap, call macb_configure_dma() which writes buffer size to hardware registers. This is important because the change_mtu codepath changes the buffer size. - Rebase onto latest net-next/main (58dd34dbd5b0) & resolve conflicts. - Link to v1: https://patch.msgid.link/20260401-macb-context-v1-0-9590c5ab7272@bootlin.com To: Théo Lebrun To: Conor Dooley To: Andrew Lunn To: "David S. Miller" To: Eric Dumazet To: Jakub Kicinski To: Paolo Abeni To: Richard Cochran To: Russell King Cc: netdev@vger.kernel.org Cc: linux-kernel@vger.kernel.org Cc: Paolo Valerio Cc: Nicolai Buchwitz Cc: Vladimir Kondratiev Cc: Gregory CLEMENT Cc: Benoît Monin Cc: Tawfik Bayouk Cc: Thomas Petazzoni Signed-off-by: Théo Lebrun --- Théo Lebrun (8): net: macb: introduce macb_context struct for buffer management net: macb: avoid macb_init_rx_buffer_size() modifying state net: macb: make `struct macb` subset reachable from macb_context struct net: macb: change caps helpers signatures net: macb: change function signatures to take contexts net: macb: introduce macb_context_alloc() helper net: macb: use context swapping in .set_ringparam() net: macb: use context swapping in .ndo_change_mtu() drivers/net/ethernet/cadence/macb.h | 115 ++- drivers/net/ethernet/cadence/macb_main.c | 1136 +++++++++++++++++++----------- drivers/net/ethernet/cadence/macb_ptp.c | 8 +- 3 files changed, 818 insertions(+), 441 deletions(-) --- base-commit: 068b5854d6cf588826bd876172652e7c1469d4e6 change-id: 20260401-macb-context-bd0caf20414d Best regards, -- Théo Lebrun