napi_build_skb() reuses per-cpu NAPI skbuff_head cache in order
to save some cycles on freeing/allocating skbuff_heads on every
new Rx or completed Tx.
ixgbe driver runs Tx completion polling cycle right before the Rx
one and uses napi_consume_skb() to feed the cache with skbuff_heads
of completed entries, so it's never empty and always warm at that
moment. Switch to the napi_build_skb() to relax mm pressure on
heavy Rx.
Signed-off-by: Alexander Lobakin <alexandr.lobakin@intel.com>
Reviewed-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>
Tested-by: Gurucharan G <gurucharanx.g@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
napi_build_skb() reuses per-cpu NAPI skbuff_head cache in order
to save some cycles on freeing/allocating skbuff_heads on every
new Rx or completed Tx.
igc driver runs Tx completion polling cycle right before the Rx
one and uses napi_consume_skb() to feed the cache with skbuff_heads
of completed entries, so it's never empty and always warm at that
moment. Switch to the napi_build_skb() to relax mm pressure on
heavy Rx.
Signed-off-by: Alexander Lobakin <alexandr.lobakin@intel.com>
Reviewed-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>
Tested-by: Nechama Kraus <nechamax.kraus@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
napi_build_skb() reuses per-cpu NAPI skbuff_head cache in order
to save some cycles on freeing/allocating skbuff_heads on every
new Rx or completed Tx.
igb driver runs Tx completion polling cycle right before the Rx
one and uses napi_consume_skb() to feed the cache with skbuff_heads
of completed entries, so it's never empty and always warm at that
moment. Switch to the napi_build_skb() to relax mm pressure on
heavy Rx.
Signed-off-by: Alexander Lobakin <alexandr.lobakin@intel.com>
Reviewed-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>
Tested-by: Gurucharan G <gurucharanx.g@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
napi_build_skb() reuses per-cpu NAPI skbuff_head cache in order
to save some cycles on freeing/allocating skbuff_heads on every
new Rx or completed Tx.
ice driver runs Tx completion polling cycle right before the Rx
one and uses napi_consume_skb() to feed the cache with skbuff_heads
of completed entries, so it's never empty and always warm at that
moment. Switch to the napi_build_skb() to relax mm pressure on
heavy Rx.
Signed-off-by: Alexander Lobakin <alexandr.lobakin@intel.com>
Reviewed-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>
Tested-by: Gurucharan G <gurucharanx.g@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
napi_build_skb() reuses per-cpu NAPI skbuff_head cache in order
to save some cycles on freeing/allocating skbuff_heads on every
new Rx or completed Tx.
iavf driver runs Tx completion polling cycle right before the Rx
one and uses napi_consume_skb() to feed the cache with skbuff_heads
of completed entries, so it's never empty and always warm at that
moment. Switch to the napi_build_skb() to relax mm pressure on
heavy Rx.
Signed-off-by: Alexander Lobakin <alexandr.lobakin@intel.com>
Reviewed-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>
Tested-by: Konrad Jankowski <konrad0.jankowski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
napi_build_skb() reuses per-cpu NAPI skbuff_head cache in order
to save some cycles on freeing/allocating skbuff_heads on every
new Rx or completed Tx.
i40e driver runs Tx completion polling cycle right before the Rx
one and uses napi_consume_skb() to feed the cache with skbuff_heads
of completed entries, so it's never empty and always warm at that
moment. Switch to the napi_build_skb() to relax mm pressure on
heavy Rx.
Signed-off-by: Alexander Lobakin <alexandr.lobakin@intel.com>
Reviewed-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>
Tested-by: Gurucharan G <gurucharanx.g@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
napi_build_skb() reuses per-cpu NAPI skbuff_head cache in order
to save some cycles on freeing/allocating skbuff_heads on every
new Rx or completed Tx element.
e1000 driver runs Tx completion polling cycle right before the Rx
one. Now that e1000 uses napi_consume_skb() to put skbuff_heads of
completed entries into the cache, it will never empty and always
warm at that moment. Switch to the napi_build_skb() to relax mm
pressure on heavy Rx and increase throughput.
Signed-off-by: Alexander Lobakin <alexandr.lobakin@intel.com>
Reviewed-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>
Tested-by: Tony Brelinski <tony.brelinski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
In order to take the best from per-cpu NAPI skbuff_head caches and
CPU cycles, let's switch from dev_kfree_skb_any(), which passes skb
back to the mm layer, to napi_consume_skb(), which feeds those
caches on non-zero budget instead (falls back to the former on 0).
Do the replacement in e1000_unmap_and_free_tx_resource(). There are
4 call sites of this function throughout the driver:
* e1000_clean_tx_ring(). Slowpath, process context, cleans the
whole Tx ring on ifdown. Use budget of 0 here;
* e1000_tx_map(). Hotpath, net Tx softirq, unmaps the buffers in
case of error. Use 0 as well;
* e1000_clean_tx_irq(). Hotpath, NAPI Tx completion polling cycle.
As the driver doesn't count completed Tx entries towards the NAPI
budget, just use the poll budget of 64 to utilize caches.
Apart from being a preparation for switching to napi_build_skb(),
this is useful on its own as well, as napi_consume_skb() flushes
skb caches by batches of 32 instead of one-at-a-time.
Signed-off-by: Alexander Lobakin <alexandr.lobakin@intel.com>
Reviewed-by: Michal Swiatkowski <michal.swiatkowski@linux.intel.com>
Tested-by: Tony Brelinski <tony.brelinski@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
The bindings for Altera and Imagination extensions are the same as for
the original Synopsys Designware Mobile Storage Host Controller.
Integrate them into Synopsys bindings to have dtschema coverage.
Signed-off-by: Krzysztof Kozlowski <krzysztof.kozlowski@canonical.com>
Link: https://lore.kernel.org/r/20211227133558.135185-3-krzysztof.kozlowski@canonical.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
We don't care in which order are "rx" and "tx" DMA names supplied.
Fixes: 4df297aaeb ("dt-bindings: mmc: Add missing properties used in examples")
Signed-off-by: David Heidelberg <david@ixit.cz>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20211224192545.74528-1-david@ixit.cz
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
The driver overrides the error codes returned by platform_get_irq() to
-ENODEV. Switch to propagating the error codes upstream.
Signed-off-by: Sergey Shtylyov <s.shtylyov@omp.ru>
Tested-by: Manuel Lauss <manuel.lauss@gmail.com
Link: https://lore.kernel.org/r/f642ef4d-6027-eb2e-0257-1c4f13911aed@omp.ru
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Drop the SD clock's base clock for an SDR104 card, as frequency is wrong when
inserting and SD2.0/SDR50 card.
Signed-off-by: Fred Ai <fred.ai@bayhubtech.com>
Acked-by: Adrian Hunter <adrian.hunter@intel.com>
Link: https://lore.kernel.org/r/20211216060824.357-2-fred.ai@bayhubtech.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
In commit 3c1a884436 ("mmc: mediatek: add support of source_cg clock")
an independent cg was introduced to avoid a hardware hang issue during
clock mode switches (subsequent commits will set that clock as optional).
When this clock is not present in device-tree, any operation is being
done on src_clk's parent (calling clk_get_parent()): to simplify this
and avoid checking for src_clk_cg presence everytime, just assign the
parent clock to src_clk_cg and remove the now useless checks.
Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://lore.kernel.org/r/20211216125748.179602-5-angelogioacchino.delregno@collabora.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
If we get an error during probe of an optional clock with function
devm_clk_get_optional(), this means that the clock was provided, but
an error occurred: this has to be escalated to the probe function
for the driver probe to fail accordingly, or unexpected hardware
behavior may happen.
Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://lore.kernel.org/r/20211216125748.179602-4-angelogioacchino.delregno@collabora.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
This driver was unconditionally enabling support for SDIO mode,
but we do have a generic "no-sdio" DT parameter that sets caps2
MMC_CAP2_NO_SDIO.
Modify the HW initialization sequence to enable SDIO support only
if it hasn't been explicitly disabled in device-tree.
Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://lore.kernel.org/r/20211216125748.179602-3-angelogioacchino.delregno@collabora.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Register fields are currently represented with open-coded bit
shifting: replace all occurrences of that to make use of kernel
provided macros BIT and GENMASK to increase human readability.
This patch provides no functional change.
Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://lore.kernel.org/r/20211216125748.179602-2-angelogioacchino.delregno@collabora.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Replace all instances of open-coded while loops for polling registers
with calls to readl_poll_timeout() and, while at it, also fix some
possible infinite loop instances.
Signed-off-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://lore.kernel.org/r/20211216125748.179602-1-angelogioacchino.delregno@collabora.com
Signed-off-by: Ulf Hansson <ulf.hansson@linaro.org>
Function _ReadLEDSetting() sets only a single variable and the only
user is ReadAdapterInfo8188EU(). Remove _ReadLEDSetting() and set the
variable in ReadAdapterInfo8188EU() directly.
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20211228101120.9120-11-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The field RSSI_test of struct odm_dm_struct is never set. It stays
at the default value 0. Remove it and remove a related if test that
is always true.
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20211228101120.9120-10-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
In this driver pDM_FatTable->FAT_State is always FAT_NORMAL_STATE.
So the check 'if (pDM_FatTable->FAT_State == FAT_TRAINING_STATE)'
is always false. Remove dead code that is executed only if that
check is true.
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20211228101120.9120-7-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The fields pre_cca_state, cur_cca_state and rssi_val_min of struct
rtl_ps are set but never used. Remove them.
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20211228101120.9120-6-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
ODM_CmnInfoInit() is never called with ODM_CMNINFO_ABILITY.
Remove that unused case from ODM_CmnInfoInit().
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20211228101120.9120-5-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The field DM_PriCCA of struct odm_dm_struct is set but never used.
Remove it and remove related dead code.
Acked-by: Martin Kaiser <martin@kaiser.cx>
Signed-off-by: Michael Straube <straube.linux@gmail.com>
Link: https://lore.kernel.org/r/20211228101120.9120-2-straube.linux@gmail.com
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove the private "test" ioctl. It copies data from user space,
this data is not used.
We can now remove a number of NULL entries at the end of the private
ioctl list.
Signed-off-by: Martin Kaiser <martin@kaiser.cx>
Link: https://lore.kernel.org/r/20211226212535.197989-10-martin@kaiser.cx
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove the private ioctls that are mapped to rtw_wx_priv_null.
rtw_wx_priv_null itself can also be removed.
Signed-off-by: Martin Kaiser <martin@kaiser.cx>
Link: https://lore.kernel.org/r/20211226212535.197989-6-martin@kaiser.cx
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove the entries of the rtw_private_args array that refer to
non-existing private ioctls.
Signed-off-by: Martin Kaiser <martin@kaiser.cx>
Link: https://lore.kernel.org/r/20211226212535.197989-2-martin@kaiser.cx
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
rfoff_reason in struct pwrctrl_priv is never set, its value remains 0.
Remove rfoff_reason, related defines and a check in rtw_led_control
that is always false.
Signed-off-by: Martin Kaiser <martin@kaiser.cx>
Link: https://lore.kernel.org/r/20211226195556.159471-22-martin@kaiser.cx
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
rtw_led_control is the only caller of SwLedControlMode1.
Pull SwLedControlMode1 into rtw_led_control.
Signed-off-by: Martin Kaiser <martin@kaiser.cx>
Link: https://lore.kernel.org/r/20211226195556.159471-21-martin@kaiser.cx
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Move BlinkingLedState updates out of if clauses where the same update
is done for all possible paths.
Signed-off-by: Martin Kaiser <martin@kaiser.cx>
Link: https://lore.kernel.org/r/20211226195556.159471-19-martin@kaiser.cx
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Remove the temporary variable bStopBlinking and check the conditions
directly in the if clauses. There's no need to save the result of
these checks.
Signed-off-by: Martin Kaiser <martin@kaiser.cx>
Link: https://lore.kernel.org/r/20211226195556.159471-18-martin@kaiser.cx
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
None of the callers sets LED_CTL_START_WPS_BOTTON. Remove the define
and related dead code.
Signed-off-by: Martin Kaiser <martin@kaiser.cx>
Link: https://lore.kernel.org/r/20211226195556.159471-17-martin@kaiser.cx
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
The LED_CTL_POWER_ON mode is not used by this driver.
Remove the define and related dead code.
Signed-off-by: Martin Kaiser <martin@kaiser.cx>
Link: https://lore.kernel.org/r/20211226195556.159471-16-martin@kaiser.cx
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Export the function that other layers use for setting the led.
Remove the function pointer and the macro to make the led control function
configurable. This driver supports only a single configuration.
Signed-off-by: Martin Kaiser <martin@kaiser.cx>
Link: https://lore.kernel.org/r/20211226195556.159471-15-martin@kaiser.cx
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
