It's redundant to prepare resources for Read and Atomic
requests by different functions. Replace them by a common
rxe_prepare_res() with different parameters. In addition,
the common rxe_prepare_res() can also be used by new Flush
and Atomic Write requests in the future.
Link: https://lore.kernel.org/r/20220705145212.12014-1-yangx.jy@fujitsu.com
Signed-off-by: Xiao Yang <yangx.jy@fujitsu.com>
Reviewed-by: Bob Pearson <rpearsonhpe@gmail.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
- preparation for new chipset support
- fixes
- VHT/HE related improvements
- ACPI SAR support
-----BEGIN PGP SIGNATURE-----
Comment: GPGTools - http://gpgtools.org
iF0EABECAB0WIQR10Rp9kadxD0kAQu/XfRQdAqdu9QUCYswM/gAKCRDXfRQdAqdu
9e8MAKDE95xMfVuG+qUHYUnUiq0US5+z1ACeO4AQu0lHl73AgmsQwRp69/0v8+c=
=OO4M
-----END PGP SIGNATURE-----
Merge tag 'mt76-for-kvalo-2022-07-11' of https://github.com/nbd168/wireless
mt76 patches for 5.20
- preparation for new chipset support
- fixes
- VHT/HE related improvements
- ACPI SAR support
[kvalo@kernel.org: fix mac80211 API changes in start_ap() and stop_ap()]
Use clk_get_rate for base_clk on resume before setting new rate.
This change ensures that the clock api returns current rate
and sets the clock to the desired rate and honors CLK_GET_NO_CACHE
attribute used by clock api.
Fixes: 97904a5985 (mmc: sdhci-brcmstb: Add ability to increase max clock rate for 72116b0)
Signed-off-by: Kamal Dasu <kdasu.kdev@gmail.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Link: https://lore.kernel.org/r/20220714174132.18541-1-kdasu.kdev@gmail.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
The function rxe_create_qp calls rxe_qp_from_init. If some error
occurs, the error handler of function rxe_qp_from_init will set
both scq and rcq to NULL.
Then rxe_create_qp calls rxe_put to handle qp. In the end,
rxe_qp_do_cleanup is called by rxe_put. rxe_qp_do_cleanup directly
accesses scq and rcq before checking them. This will cause
null-ptr-deref error.
The call graph is as below:
rxe_create_qp {
...
rxe_qp_from_init {
...
err1:
...
qp->rcq = NULL; <---rcq is set to NULL
qp->scq = NULL; <---scq is set to NULL
...
}
qp_init:
rxe_put{
...
rxe_qp_do_cleanup {
...
atomic_dec(&qp->scq->num_wq); <--- scq is accessed
...
atomic_dec(&qp->rcq->num_wq); <--- rcq is accessed
}
}
Fixes: 4703b4f0d9 ("RDMA/rxe: Enforce IBA C11-17")
Link: https://lore.kernel.org/r/20220705225414.315478-1-yanjun.zhu@linux.dev
Signed-off-by: Zhu Yanjun <yanjun.zhu@linux.dev>
Reviewed-by: Bob Pearson <rpearsonhpe@gmail.com>
Reviewed-by: Md Haris Iqbal <haris.iqbal@ionos.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
This reverts three commits:
1. Revert "mmc: sdhci-msm: Add compatible string check for sdx65"
This reverts commit 953706844f.
2. Revert "mmc: sdhci-msm: Add compatible string check for sm8150"
This reverts commit 5acd6adb65.
3. Revert "mmc: sdhci-msm: Add SoC specific compatibles"
This reverts commit 466614a976.
The oldest commit 466614a976 ("mmc: sdhci-msm: Add SoC specific
compatibles") did not specify what benefits such multiple compatibles
bring, therefore assume there is none. On the other hand such approach
brings a lot of churn to driver maintenance by expecting commit for
every new compatible, even though it is already covered by the fallback.
There is really no sense in duplicating of_device_id for each
variant, which is already covered by generic compatible fallback
qcom,sdhci-msm-v4 or qcom,sdhci-msm-v5.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Link: https://lore.kernel.org/r/20220714091042.22287-4-krzysztof.kozlowski@linaro.org
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Add a MSM8998-specific SDCC compatible, because using only a generic
qcom,sdhci-msm-v4 fallback is deprecated.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@linaro.org>
Reviewed-by: Douglas Anderson <dianders@chromium.org>
Link: https://lore.kernel.org/r/20220714091042.22287-2-krzysztof.kozlowski@linaro.org
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Send DPCD DP_SET_POWER_D0 command to the monitor in .atomic_enable
callback. Without this command, some monitors won't show up again after
changing the resolution.
Fixes: 46ca7da7f1 ("drm/bridge: it6505: Send DPCD SET_POWER to downstream")
Signed-off-by: Pin-Yen Lin <treapking@chromium.org>
Reviewed-by: Allen Chen <allen.chen@ite.com.tw>
Fixes: 46ca7da7f1 ("drm/bridge: it6505: Send DPCD SET_POWER to downstream")
Signed-off-by: Robert Foss <robert.foss@linaro.org>
Link: https://patchwork.freedesktop.org/patch/msgid/20220714173715.v2.1.I85af54e9ceda74ec69f661852825845f983fc343@changeid
Kuniyuki Iwashima says:
====================
sysctl: Fix data-races around ipv4_net_table (Round 3).
This series fixes data-races around 21 knobs after
igmp_link_local_mcast_reports in ipv4_net_table.
These 4 knobs are skipped because they are safe.
- tcp_congestion_control: Safe with RCU and xchg().
- tcp_available_congestion_control: Read only.
- tcp_allowed_congestion_control: Safe with RCU and spinlock().
- tcp_fastopen_key: Safe with RCU and xchg()
So, round 4 will start with fib_multipath_use_neigh.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_fastopen_blackhole_timeout, it can be changed
concurrently. Thus, we need to add READ_ONCE() to its readers.
Fixes: cf1ef3f071 ("net/tcp_fastopen: Disable active side TFO in certain scenarios")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_fastopen, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its readers.
Fixes: 2100c8d2d9 ("net-tcp: Fast Open base")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Acked-by: Yuchung Cheng <ycheng@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_max_syn_backlog, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its readers.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_tw_reuse, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its reader.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_notsent_lowat, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its reader.
Fixes: c9bee3b7fd ("tcp: TCP_NOTSENT_LOWAT socket option")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading these sysctl knobs, they can be changed concurrently.
Thus, we need to add READ_ONCE() to their readers.
- tcp_retries1
- tcp_retries2
- tcp_orphan_retries
- tcp_fin_timeout
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_reordering, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its readers.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_migrate_req, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its readers.
Fixes: f9ac779f88 ("net: Introduce net.ipv4.tcp_migrate_req.")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_syncookies, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its readers.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_syn(ack)?_retries, they can be changed
concurrently. Thus, we need to add READ_ONCE() to their readers.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_tcp_keepalive_(time|probes|intvl), they can be changed
concurrently. Thus, we need to add READ_ONCE() to their readers.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_igmp_qrv, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its readers.
This test can be packed into a helper, so such changes will be in the
follow-up series after net is merged into net-next.
qrv ?: READ_ONCE(net->ipv4.sysctl_igmp_qrv);
Fixes: a9fe8e2994 ("ipv4: implement igmp_qrv sysctl to tune igmp robustness variable")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_igmp_max_msf, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its readers.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_igmp_max_memberships, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its reader.
Fixes: 1da177e4c3 ("Linux-2.6.12-rc2")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
While reading sysctl_igmp_llm_reports, it can be changed concurrently.
Thus, we need to add READ_ONCE() to its readers.
This test can be packed into a helper, so such changes will be in the
follow-up series after net is merged into net-next.
if (ipv4_is_local_multicast(pmc->multiaddr) &&
!READ_ONCE(net->ipv4.sysctl_igmp_llm_reports))
Fixes: df2cf4a78e ("IGMP: Inhibit reports for local multicast groups")
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If siw_recv_mpa_rr returns -EAGAIN, it means that the MPA reply hasn't
been received completely, and should not report IW_CM_EVENT_CONNECT_REPLY
in this case. This may trigger a call trace in iw_cm. A simple way to
trigger this:
server: ib_send_lat
client: ib_send_lat -R <server_ip>
The call trace looks like this:
kernel BUG at drivers/infiniband/core/iwcm.c:894!
invalid opcode: 0000 [#1] PREEMPT SMP NOPTI
<...>
Workqueue: iw_cm_wq cm_work_handler [iw_cm]
Call Trace:
<TASK>
cm_work_handler+0x1dd/0x370 [iw_cm]
process_one_work+0x1e2/0x3b0
worker_thread+0x49/0x2e0
? rescuer_thread+0x370/0x370
kthread+0xe5/0x110
? kthread_complete_and_exit+0x20/0x20
ret_from_fork+0x1f/0x30
</TASK>
Fixes: 6c52fdc244 ("rdma/siw: connection management")
Link: https://lore.kernel.org/r/dae34b5fd5c2ea2bd9744812c1d2653a34a94c67.1657706960.git.chengyou@linux.alibaba.com
Signed-off-by: Cheng Xu <chengyou@linux.alibaba.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Allow the guest to know how much it is ballooned by the host.
It is useful when debugging out of memory conditions.
When host gets back memory from the guest it is accounted
as used memory in the guest but the guest have no way to know
how much it is actually ballooned.
Expose current state, flags and max possible memory to the guest.
While at it - fix a 10+ years old typo.
Signed-off-by: Alexander Atanasov <alexander.atanasov@virtuozzo.com>
Reviewed-by: Michael Kelley <mikelley@microsoft.com>
Link: https://lore.kernel.org/r/20220711181825.52318-1-alexander.atanasov@virtuozzo.com
Signed-off-by: Wei Liu <wei.liu@kernel.org>
Merge "Sunplus SP7021 SoC support" from Qin Jian, applied
as patches:
"Sunplus SP7021 is an ARM Cortex A7 (4 cores) based SoC. It integrates many
peripherals (ex: UART, I2C, SPI, SDIO, eMMC, USB, SD card and etc.) into a
single chip. It is designed for industrial control.
SP7021 consists of two chips (dies) in a package. One is called C-chip
(computing chip). It is a 4-core ARM Cortex A7 CPU. It adopts high-level
process (22 nm) for high performance computing. The other is called P-
chip (peripheral chip). It has many peripherals and an ARM A926 added
especially for real-time control. P-chip is made for customers. It adopts
low-level process (ex: 0.11 um) to reduce cost.
Refer to (for documentations):
https://sunplus-tibbo.atlassian.net/wiki/spaces/doc/overview
Refer to (applications):
https://tibbo.com/store/plus1.html
Refer to (applications):
http://www.sinovoip.com.cn/ecp_view.asp?id=586"
I left out the clock controller driver, which is still not fully
reviewed. There have been 20 revisions of the platform code, and
everything else looks good enough, so I hope it's we can simplify
the process by separating it out again.
Link: https://lore.kernel.org/linux-arm-kernel/0c10fa4ccf3da5d92784b9bbd4177d1d2f1d62a0.1656396767.git.qinjian@cqplus1.com/
* sunplus/newsoc:
ARM: dts: Add Sunplus SP7021-Demo-V3 board device tree
ARM: sp7021_defconfig: Add Sunplus SP7021 defconfig
ARM: sunplus: Add initial support for Sunplus SP7021 SoC
irqchip: Add Sunplus SP7021 interrupt controller driver
dt-bindings: interrupt-controller: Add bindings for SP7021 interrupt controller
dt-bindings: clock: Add bindings for SP7021 clock driver
reset: Add Sunplus SP7021 reset driver
dt-bindings: reset: Add bindings for SP7021 reset driver
dt-bindings: arm: sunplus: Add bindings for Sunplus SP7021 SoC boards
Since ECC memory maintains a memory system immune to single-bit errors,
add support for correcting the 1bit-ECC error, which prevents a 1bit-ECC
error become an uncorrected type error. When a 1bit-ECC error happens in
the internal ram of the ROCE engine, such as the QPC table, as a 1bit-ECC
error caused by reading, the ROCE engine only corrects those 1bit ECC
errors by writing.
Link: https://lore.kernel.org/r/20220714134353.16700-6-liangwenpeng@huawei.com
Signed-off-by: Haoyue Xu <xuhaoyue1@hisilicon.com>
Signed-off-by: Wenpeng Liang <liangwenpeng@huawei.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Use a single function to handle the same kind of abnormal interrupts.
Link: https://lore.kernel.org/r/20220714134353.16700-5-liangwenpeng@huawei.com
Signed-off-by: Haoyue Xu <xuhaoyue1@hisilicon.com>
Signed-off-by: Wenpeng Liang <liangwenpeng@huawei.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
The driver will clear all the interrupts in the same area
when the driver handles the interrupt of type AEQ overflow.
It should only set the interrupt status bit of type AEQ overflow.
Fixes: a5073d6054 ("RDMA/hns: Add eq support of hip08")
Link: https://lore.kernel.org/r/20220714134353.16700-4-liangwenpeng@huawei.com
Signed-off-by: Haoyue Xu <xuhaoyue1@hisilicon.com>
Signed-off-by: Wenpeng Liang <liangwenpeng@huawei.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
The type of return value of the interrupt handler should be irqreturn_t.
Link: https://lore.kernel.org/r/20220714134353.16700-3-liangwenpeng@huawei.com
Signed-off-by: Haoyue Xu <xuhaoyue1@hisilicon.com>
Signed-off-by: Wenpeng Liang <liangwenpeng@huawei.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
The HNS NIC driver receives and handles the abnormal interrupt of the RAS
type generated by ROCEE, and the HNS RDMA driver does not need to handle
this type of interrupt. Therefore, delete unused codes in the HNS RDMA
driver.
Link: https://lore.kernel.org/r/20220714134353.16700-2-liangwenpeng@huawei.com
Signed-off-by: Haoyue Xu <xuhaoyue1@hisilicon.com>
Signed-off-by: Wenpeng Liang <liangwenpeng@huawei.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
__qedr_alloc_mr() allocates a memory chunk for "mr->info.pbl_table" with
init_mr_info(). When rdma_alloc_tid() and rdma_register_tid() fail, "mr"
is released while "mr->info.pbl_table" is not released, which will lead
to a memory leak.
We should release the "mr->info.pbl_table" with qedr_free_pbl() when error
occurs to fix the memory leak.
Fixes: e0290cce6a ("qedr: Add support for memory registeration verbs")
Link: https://lore.kernel.org/r/20220714061505.2342759-1-niejianglei2021@163.com
Signed-off-by: Jianglei Nie <niejianglei2021@163.com>
Acked-by: Michal Kalderon <michal.kalderon@marvell.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
It was observed that by allowing pinctrl_amd to be loaded
later in the boot process that interrupts sent to the GPIO
controller early in the boot are not serviced. The kernel treats
these as a spurious IRQ and disables the IRQ.
This problem was exacerbated because it happened on a system with
an encrypted partition so the kernel object was not accesssible for
an extended period of time while waiting for a passphrase.
To avoid this situation from occurring, stop allowing pinctrl-amd
from being built as a module and instead require it to be built-in
or disabled.
Reported-by: madcatx@atlas.cz
Suggested-by: jwrdegoede@fedoraproject.org
Link: https://bugzilla.kernel.org/show_bug.cgi?id=216230
Signed-off-by: Mario Limonciello <mario.limonciello@amd.com>
Acked-by: Hans de Goede <hdegoede@redhat.com>
Link: https://lore.kernel.org/r/20220713175950.964-1-mario.limonciello@amd.com
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Adjusted as per packet processor documentation.
This allows to properly match 'indev' for clsact rules.
Fixes: 47327e198d ("net: prestera: acl: migrate to new vTCAM api")
Signed-off-by: Maksym Glubokiy <maksym.glubokiy@plvision.eu>
Signed-off-by: David S. Miller <davem@davemloft.net>
The LED controller node is disabled because the leds-turris-omnia driver
does not support setting the LED blinking to be controlled by the MCU.
The patches for that have now been sent [1], so let's enable the node.
[1] https://lore.kernel.org/linux-leds/20220704105955.15474-1-kabel@kernel.org/T/
Signed-off-by: Marek Behún <kabel@kernel.org>
Signed-off-by: Gregory CLEMENT <gregory.clement@bootlin.com>
The marvell PHY driver changes the LED[0] pin function to "On - 1000
Mbps Link, Off - Else".
Turris Omnia expects that the function is "On - Link, Blink - Activity,
Off - No link".
Use the `marvell,reg-init` DT property to change the function.
In the future, once netdev trigger will support HW offloading, we will
be able to have this configured via the combination of PHY driver and
leds-turris-omnia driver.
Signed-off-by: Marek Behún <kabel@kernel.org>
Signed-off-by: Gregory CLEMENT <gregory.clement@bootlin.com>
Both hfi1 and UML depend on x86_64, this can trigger build errors.
This driver must depends on !UML because it accesses x86_64
features that are not supported by UML.
Link: https://lore.kernel.org/r/165755127879.2996325.5668395672492732376.stgit@awfm-02.cornelisnetworks.com
Signed-off-by: Ehab Ababneh <ehab.ababneh@cornelisnetworks.com>
Signed-off-by: Dennis Dalessandro <dennis.dalessandro@cornelisnetworks.com>
Signed-off-by: Leon Romanovsky <leon@kernel.org>
Socket destruction flow and tls_device_down function sync against each
other using tls_device_lock and the context refcount, to guarantee the
device resources are freed via tls_dev_del() by the end of
tls_device_down.
In the following unfortunate flow, this won't happen:
- refcount is decreased to zero in tls_device_sk_destruct.
- tls_device_down starts, skips the context as refcount is zero, going
all the way until it flushes the gc work, and returns without freeing
the device resources.
- only then, tls_device_queue_ctx_destruction is called, queues the gc
work and frees the context's device resources.
Solve it by decreasing the refcount in the socket's destruction flow
under the tls_device_lock, for perfect synchronization. This does not
slow down the common likely destructor flow, in which both the refcount
is decreased and the spinlock is acquired, anyway.
Fixes: e8f6979981 ("net/tls: Add generic NIC offload infrastructure")
Reviewed-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Signed-off-by: Tariq Toukan <tariqt@nvidia.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
- The code relies on rc_pci_fixup being called, which only happens
when CONFIG_PCI_QUIRKS is enabled, so add that to Kconfig. Omitting
this causes a booting failure with a non-obvious cause.
- Update rc_pci_fixup to set the class properly, copying the
more modern style from other places
- Correct the rc_pci_fixup comment
This patch just re-applies commit 1dc831bf53 ("ARM: Kirkwood: Update
PCI-E fixup") for all other Marvell ARM platforms which have same buggy
PCIe controller and do not use pci-mvebu.c controller driver yet.
Long-term goal for these Marvell ARM platforms should be conversion to
pci-mvebu.c controller driver and removal of these fixups in arch code.
Signed-off-by: Pali Rohár <pali@kernel.org>
Cc: Jason Gunthorpe <jgg@nvidia.com>
Cc: stable@vger.kernel.org
Signed-off-by: Gregory CLEMENT <gregory.clement@bootlin.com>
Use netif_napi_add_tx() for NAPI in Tx direction instead of the regular
netif_napi_add() function.
Signed-off-by: Sieng-Piaw Liew <liew.s.piaw@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
This patch removes the of_match_ptr() pointer when dereferencing the
ksz_dt_ids which produce the unused variable warning.
Reported-by: kernel test robot <lkp@intel.com>
Suggested-by: Arnd Bergmann <arnd@kernel.org>
Signed-off-by: Arun Ramadoss <arun.ramadoss@microchip.com>
Reviewed-by: Arnd Bergmann <arnd@arndb.de>
Signed-off-by: David S. Miller <davem@davemloft.net>
Jakub Kicinski says:
====================
tls: rx: avoid skb_cow_data()
TLS calls skb_cow_data() on the skb it received from strparser
whenever it needs to hold onto the skb with the decrypted data.
(The alternative being decrypting directly to a user space buffer
in whic case the input skb doesn't get modified or used after.)
TLS needs the decrypted skb:
- almost always with TLS 1.3 (unless the new NoPad is enabled);
- when user space buffer is too small to fit the record;
- when BPF sockmap is enabled.
Most of the time the skb we get out of strparser is a clone of
a 64kB data unit coalsced by GRO. To make things worse skb_cow_data()
tries to output a linear skb and allocates it with GFP_ATOMIC.
This occasionally fails even under moderate memory pressure.
This patch set rejigs the TLS Rx so that we don't expect decryption
in place. The decryption handlers return an skb which may or may not
be the skb from strparser. For TLS 1.3 this results in a 20-30%
performance improvement without NoPad enabled.
v2: rebase after 3d8c51b25a ("net/tls: Check for errors in tls_device_init")
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
We currently CoW Rx skbs whenever we can't decrypt to a user
space buffer. The skbs can be enormous (64kB) and CoW does
a linear alloc which has a strong chance of failing under
memory pressure. Or even without, skb_cow_data() assumes
GFP_ATOMIC.
Allocate a new frag'd skb and decrypt into it. We finally
take advantage of the decrypted skb getting returned via
darg.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
The "zero-copy" path in SW TLS will engage either for no skbs or
for all but last. If the recvmsg parameters are right and the
socket can do ZC we'll ZC until the iterator can't fit a full
record at which point we'll decrypt one more record and copy
over the necessary bits to fill up the request.
The only reason we hold onto the ZC skbs which went thru the async
path until the end of recvmsg() is to count bytes. We need an accurate
count of zc'ed bytes so that we can calculate how much of the non-zc'd
data to copy. To allow freeing input skbs on the ZC path count only
how much of the list we'll need to consume.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Async crypto currently benefits from the fact that we decrypt
in place. When we allow input and output to be different skbs
we will have to hang onto the input while we move to the next
record. Clone the inputs and keep them on a list.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Async crypto TLS Rx currently waits for crypto to be done
in order to strip the TLS header and tailer. Simplify
the code by moving the pointers immediately, since only
TLS 1.2 is supported here there is no message padding.
This simplifies the decryption into a new skb in the next
patch as we don't have to worry about input vs output
skb in the decrypt_done() handler any more.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
