From: Sandeep Haemoon <sandeeph@loljews.com>
To: netdev@vger.kernel.org
Cc: Felix Fietkau <nbd@nbd.name>,
Lorenzo Bianconi <lorenzo@kernel.org>,
Andrew Lunn <andrew+netdev@lunn.ch>,
David S. Miller <davem@davemloft.net>,
Eric Dumazet <edumazet@kernel.org>,
Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
Matthias Brugger <matthias.bgg@gmail.com>,
AngeloGioacchino Del Regno
<angelogioacchino.delregno@collabora.com>,
Russell King <linux@armlinux.org.uk>,
Daniel Golle <daniel@makrotopia.org>,
Mason Chang <mason-cw.chang@mediatek.com>,
Frank Wunderlich <frank-w@public-files.de>,
linux-mediatek@lists.infradead.org,
linux-arm-kernel@lists.infradead.org,
linux-kernel@vger.kernel.org
Subject: [PATCH net-next v8 0/3] net: ethernet: mtk_eth_soc: 9K jumbo frame support
Date: Mon, 23 Nov 2026 01:13:31 +0200 [thread overview]
Message-ID: <patch-51fa5ed3bc801edb3f29b61ca7888382@loljews.com> (raw)
This series makes the MediaTek SoC ethernet driver support 9K jumbo
frames on the 10G XGMII MACs (MT7988). The driver advertises a
2030-byte max_mtu and programs the MAC max-rx field for at most 2048
bytes, so MTU > 2K is rejected and 9K frames are not usable.
Patch 1 fixes a pre-existing 36-bit DMA bug in mtk_poll_rx() that the
review of v3 found: a descriptor released without a new mapping is
handed back with the previous buffer's high address bits. It is
independent of the rest of the series and carries a Fixes: tag for the
commit that introduced the bug, so the maintainers can route it to
net directly if they prefer.
Patch 2 gates the page_pool on the ring's descriptor length: the pool
hands out order-0 pages and maps at most MTK_PP_MAX_BUF_SIZE, so a
ring whose descriptor length is larger has to use the
netdev_alloc_frag()/mtk_max_buf_alloc() path.
Patch 3 adds the MTK_NETSYS_RX_9K capability, advertises a 9194-byte
max_mtu on the XGMII MACs (from the DT phy_mode at probe, so DSA can
raise the conduit MTU before the netdev is opened), programs
XMAC_RX_CFG2 and the PDMA RX SDL, derives the shared RX buffer length
from the largest MTU (802.1Q headroom included) and sizes the RX rings
from it, keeping the never-polled QDMA ring at its 2K buffers. The
MAC max-rx limits stay in sync with that shared length, including
GMAC1's MCR jumbo field. A runtime MTU change that needs larger
buffers is rejected with -EBUSY while DMA is running; shrinking is
allowed. GMAC1, the built-in switch port, keeps its MCR programming
(bucket plus the jumbo field in 1024-byte units), written after the
MAC_MCR_FORCE_LINK_DOWN reset so it survives every mac_config();
whether the internal switch path enforces that field at all is a
hardware property the commit message documents.
This work is a forward-port of the MediaTek downstream jumbo work:
Mason Chang's "change default rx buffer length" and "add 9k jumbo
frame support" (MediaTek OpenWrt feed 999-eth-30/31, carried in
frank-w/BPI-Router-Linux), plus Frank Wunderlich's "try to fix crash
the right way". The runtime reallocation of Mason's "dynamic rx
buffer adjustment" (feed 999-eth-33) is not included and will be sent
separately. Attribution is prose plus Link: tags: Co-developed-by
requires the original authors' own Signed-off-by lines, which we do
not have, and the port diverges from the downstream implementation
(2K buffer floor, per-ring pool gate, the -EBUSY rule, the XDP checks
and the truesize fix are not in the downstream patches).
The equivalent change was tested on a BPI-R4 Pro 8X (MT7988, 10G
lan6 / MxL86252) on the 6.18.60 backport: with both ends at MTU 9000,
8K ICMP passes and iperf3 with MSS 8900 reaches ~9.87 Gbit/s (vs
~2.57 Gbit/s at MSS 1460); the DSA conduit is raised to 9004 at boot.
This series is compile-tested with CONFIG_NET_MEDIATEK_SOC=y and
COMPILE_TEST=y; no hardware run of this exact tree yet.
With the 1 KiB rounding the RX buffer length is order-1 (8 KiB) for
rx_buf_len 4096..7168 and order-2 (16 KiB) from rx_buf_len 8192 up;
at the 9K ring size the 16 KiB buffers pin around 64 MiB while DMA is
up and make the refill a high-order GFP_ATOMIC allocation. That cost
is inherent to the downstream design; skb->truesize now reports the
allocation that was actually made.
Changes in v8 (the v7 review and the published reference manual):
- the 2K floor is now applied before the 802.1Q headroom is added, so
MTU 2027..2030 stays at 2K and neither the SGMII/2500base-x MACs on
MT7988 nor a DSA user port at 2023..2026 fails a runtime change with
-EBUSY. A tagged maximum-size frame in that class is still subject
to the historical 2K limit;
- MTU changes, mtk_dma_init() and mtk_hw_init() now program every MAC
with the shared ring length instead of only the device being changed,
so no MAC keeps a limit larger than the buffers after a shrink or a
partial reopen. MACs whose interface mode is not known yet are
skipped and left to their own mac_config();
- the XMAC max-RX limit is written again after the XMAC logic reset that
mtk_xgdm_mac_link_up() pulses: the published MT7988A reference manual
gives RX_CFG2 a reset value of 0x800 (2048 bytes), so a reset would
otherwise cap the MAC at 2K;
- the PDMA SDL write is now described from the MT7988A reference
manual: ADMA_RX_CFG[31:16] is the "ADMA LRO free buffer SDL", a
buffer capacity, not the RX frame-length limit; the mask is widened
to the documented GENMASK(31, 16). The write itself is unchanged and
matches the vendor driver;
- the MCR jumbo field keeps the vendor encoding; the reference manual
defines only 4'h2 (2 Kbytes) for it and marks the rest reserved, but
the vendor driver programs the length and jumbo passes on its
reference hardware;
- the commit messages now describe the over-long descriptor drop and the
exported rx_length_errors update, the all-MAC limit pass and the
link-up rewrite; the page_pool comment names mtk_max_buf_alloc() after
the rename; the new buffer-size helper is not inline;
- 1/3 is retitled to "reset dma_addr for every RX descriptor";
- 2/3 no longer carries Simon Horman's Reviewed-by: the predicate it was
given for changed in v3, before the tag was carried over.
The 36-bit DMA fix in 1/3 was found by review of an earlier revision,
not in production, and has not been triggered at runtime here: it needs
MTK_36BIT_DMA, DRAM above 4 GiB and one of the early release paths in
mtk_poll_rx().
Tested on a BPI-R4 Pro 8X (MT7988, 10G port) on a 6.18.60 backport:
8K ICMP passes and iperf3 with MSS 8900 at MTU 9000 reaches about
9.87 Gbit/s. The 36-bit fix itself was not exercised there, and this
series is compile-tested with CONFIG_NET_MEDIATEK_SOC=y and
COMPILE_TEST=y.
Changes in v7:
- respin reason: patchwork registered the v6 posting as superseded by the
v5 posting that went out ten minutes before it, so the series that
actually applies was hidden behind one that does not. This repost
restores a sane version order; the content below also folds in the v6
review.
- only the SoCs with MTK_NETSYS_RX_9K grow the shared RX rings with the
MTU. Everywhere else the buffers stay at 2K, so the 2030-byte maximum
those SoCs advertise keeps working at runtime and the -EBUSY rule no
longer applies to them.
- the XDP admission checks now test the same descriptor length the
page_pool gate uses, instead of the raw buffer length.
- mtk_get_stats64() adds the software rx_length_errors counter, so the
new over-long descriptor drop is visible in the exported statistics.
- 2/3 names mtk_max_lro_buf_alloc(), the allocator that exists at that
commit.
- 1/3 gives the commits in the right order and states that the change
depends on the earlier SER fix it cites; the 3/3 message uses the
real 32 MiB figure for the single polled ring.
Changes in v6:
- rebase onto current net-next (a76e8bf4be) so the series applies
again: net-next converted mtk_rx_alloc() to kzalloc_objs(), which
the hunk context depended on. No functional change.
Changes in v5 (the v4 review, in full):
- 802.1Q headroom (Medium): the shared RX buffer length now includes
VLAN_HLEN before the 1 KiB round-up, and the advertised max_mtu is
lowered by the same four bytes to 9194, so a tagged maximum-size
frame (9194 + 18 + 4) still fits in rx_buf_len 9216 and the XMAC
9216 cap. XMAC_RX_CFG2 and the ring buffers therefore cover a
tagged frame at every MTU, including the maximum.
- GMAC1 MCR jumbo field (Medium): the vendor driver has a second MCR
max-rx field we had not ported. mtk_set_mcr_max_rx() now clears
and programs the jumbo field (bits 31:28, 1024-byte units,
MAC_MCR_MAX_RX_JUMBO) for GMAC1, so the driver can program the same
shared length for the switch port as for the XGMII MACs. The new
field handling is limited to the MTK_NETSYS_RX_9K SoCs; every other
SoC keeps its previous MCR programming bit for bit.
- QDMA ring sizing (Medium): the helpers now take the buffer length
instead of struct mtk_eth, and mtk_rx_alloc() sizes the
never-polled QDMA RX ring at the fixed 2K length instead of the
jumbo length, so it no longer pins ~32 MiB of high-order pages.
HW-LRO rings keep their 12288-byte sizing.
- pktlen bound (Medium): mtk_poll_rx() drops descriptors whose PLEN0
exceeds ring->buf_size, with rx_length_errors, before the buffer
can be passed up, so the larger build_skb() size can no longer hand
stale memory to the stack if the engine over-reports the length.
- PDMA SDL encoding (Medium): kept as in v4; the comment and commit
message now say explicitly that the GENMASK(30,16) encoding (frame
length offset by MTK_PDMA_LRO_SDL) is the vendor's and not publicly
documented, and that a datasheet confirmation would be welcome.
- XDP detach (Medium): the pool-size check now runs only when a
program is attached, so a detach always succeeds. The shared
eth->prog weakness the review also points out is pre-existing and
is left to a separate fix.
- 3/3 message (Low): the buffer-size math is corrected (order-1 for
rx_buf_len 4096..7168, order-2 from 8192) and the message now says
that the build_skb()/truesize change also covers the HW-LRO rings
on mt2701, mt7622, mt7623 and mt7629.
- 1/3 subject and message (Low): retitled to describe the fix ("fix
the RX descriptor address on the release path") and the message now
cites the commit that introduced the release-path check and the one
that later changed it, and states the fix's dependency on the latter.
The pre-existing High finding of the review (a failure while filling
an RX ring leaks the partially filled buffers and DMA mappings
because ring->dma_size is still 0) is not part of this series: it is
in the untouched upstream code and will be sent as a separate fix that
sets ring->dma_size before the fill loop.
Simon Horman's Reviewed-by is kept on patch 2 only: patches 1 and 3
are new or changed materially since his review of v1.
Previous versions:
https://lore.kernel.org/all/patch-7d2c457e68de6b53b7348d1bdbeadcec@loljews.com/ (v7)
https://lore.kernel.org/all/patch-9e765e74d06edbba3fd85a0cd49c4bb5@loljews.com/ (v6)
https://lore.kernel.org/all/patch-f1e62a87eb0f914e1ea52a395f465998@loljews.com/ (v5)
https://lore.kernel.org/all/patch-fc08a7ef40667bf3017366e452e6d3c0@loljews.com/ (v4)
https://lore.kernel.org/all/patch-aad4cfa8d9522d1f7dbc9893ea12e813@loljews.com/ (v3)
https://lore.kernel.org/all/patch-2793186c653124c4c0b3d463bde2ea3a@loljews.com/ (v2)
https://lore.kernel.org/all/patch-4f863a941889f0bceca1c082f2d79624@loljews.com/ (v1)
Sandeep Haemoon (3):
net: ethernet: mtk_eth_soc: reset dma_addr for every RX descriptor
net: ethernet: mtk_eth_soc: skip page_pool for RX rings whose buffers
do not fit
net: ethernet: mtk_eth_soc: add 9K jumbo frame support
drivers/net/ethernet/mediatek/mtk_eth_soc.c | 398 +++++++++++++++++---
drivers/net/ethernet/mediatek/mtk_eth_soc.h | 16 +-
2 files changed, 356 insertions(+), 58 deletions(-)
base-commit: a76e8bf4be552eed20801186188e5606365870a0
--
2.53.0
next reply other threads:[~2026-11-22 23:13 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2026-11-22 23:13 Sandeep Haemoon [this message]
2026-11-22 23:13 ` [PATCH net-next v8 2/3] net: ethernet: mtk_eth_soc: skip page_pool Sandeep Haemoon
2026-11-22 23:13 ` [PATCH net-next v8 1/3] net: ethernet: mtk_eth_soc: reset dma_addr Sandeep Haemoon
2026-11-22 23:13 ` [PATCH net-next v8 3/3] net: ethernet: mtk_eth_soc: add 9K jumbo Sandeep Haemoon
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=patch-51fa5ed3bc801edb3f29b61ca7888382@loljews.com \
--to=sandeeph@loljews.com \
--cc=andrew+netdev@lunn.ch \
--cc=angelogioacchino.delregno@collabora.com \
--cc=daniel@makrotopia.org \
--cc=davem@davemloft.net \
--cc=edumazet@kernel.org \
--cc=frank-w@public-files.de \
--cc=kuba@kernel.org \
--cc=linux-arm-kernel@lists.infradead.org \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-mediatek@lists.infradead.org \
--cc=linux@armlinux.org.uk \
--cc=lorenzo@kernel.org \
--cc=mason-cw.chang@mediatek.com \
--cc=matthias.bgg@gmail.com \
--cc=nbd@nbd.name \
--cc=netdev@vger.kernel.org \
--cc=pabeni@redhat.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