Sean Wang [Fri, 1 May 2020 10:36:14 +0000 (12:36 +0200)]
mt76: mt7663u: introduce suspend/resume to mt7663u
Tested on Chromebok by "echo mem to /sys/power/state" to suspend and
then waked up by keyboard keystrokes to resume system.
Signed-off-by: Sean Wang <sean.wang@mediatek.com> Co-developed-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Introduce the mt7615_mcu_set_hif_suspend mcu command, which is usually
used to configure the interface including PCIe, USB or SDIO to the right
state during operation suspend / resume.
Signed-off-by: Sean Wang <sean.wang@mediatek.com> Co-developed-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Fri, 1 May 2020 10:36:11 +0000 (12:36 +0200)]
mt76: mt7615: introduce support for hardware beacon filter
Introduce support for hw beacon filter if available in the firmware
Co-developed-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Thu, 30 Apr 2020 07:59:31 +0000 (09:59 +0200)]
mt76: mt7615: do not mark sched_scan disabled in mt7615_scan_work
For the moment offload firmware supports just one entry in the scheduled
scan plan and so it runs till it is disabled by the userspace.
Do not mark the hw scheduled scan as disabled in mt7615_scan_work
after receiving a scan result
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Felix Fietkau [Wed, 29 Apr 2020 12:34:23 +0000 (14:34 +0200)]
mt76: mt7615: fix getting maximum tx power from eeprom
On top of the EEPROM target power, each rate can also has a power offset.
On many devices, this power offset is used to boost the tx power of lower
rates. Take this into account when parsing rate power.
The assumption here is, that the first rate (OFDM 6M or CCK 1M) has the
highest tx power
Lorenzo Bianconi [Tue, 28 Apr 2020 13:34:09 +0000 (15:34 +0200)]
mt76: mt7663: add the possibility to load firmware v2
mt7663 firmware v2 is used for embedded devices since it has more completed
features in AP mode.
Add the capability to specify which firmware load first (v3 or v2)
using prefer_offload_fw kernel parameter and fallback to the other one
if the selected firmware fails to load
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Fri, 24 Apr 2020 12:51:29 +0000 (14:51 +0200)]
mt76: mt7615: fix event report in mt7615_mcu_bss_event
Currently mt7663 devices do not support DBDC so fw events have no info
about it. Fix mt7615_mcu_bss_event that wrongly use bss_idx as DBDC
band_idx while it is vif index.
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Sean Wang [Thu, 23 Apr 2020 13:47:54 +0000 (15:47 +0200)]
mt76: mt7663: fix up BMC entry indicated to unicmd firmware
BMC entry for MT7663 unicmd firmware should be a broadcast/multicast entry,
not a unicast entry, that is GTK rekey offload would rely on.
Fixes: 138860679b2a ("mt76: mt7615: add more uni mcu commands") Cc: Soul Huang <Soul.Huang@mediatek.com> Suggested-by: YF Luo <Yf.Luo@mediatek.com> Co-developed-by: Ryder Lee <ryder.lee@mediatek.com> Signed-off-by: Ryder Lee <ryder.lee@mediatek.com> Co-developed-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Wed, 22 Apr 2020 11:07:44 +0000 (13:07 +0200)]
mt76: mt7615: move mcu bss upload before creating the sta
Run mt7615_mcu_add_bss_info routine before mt7615_mcu_sta_add since
the firmware requires the bss is created before the relative sta
Tested-by: Sean Wang <sean.wang@mediatek.com> Suggested-by: YF Luo <yf.luo@mediatek.com> Suggested-by: Lucy Hsu <lucy.hsu@mediatek.com> Co-developed-by: Soul Huang <soul.huang@mediatek.com> Signed-off-by: Soul Huang <soul.huang@mediatek.com> Co-developed-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Felix Fietkau [Wed, 22 Apr 2020 11:15:06 +0000 (13:15 +0200)]
mt76: mt7603: fix tx status rate index calculation
A switch from one rate index to the next only happens when tx count from
the current slot is greater than MT7615_RATE_RETRY, which is 1 has to be
subtracted from count, instead of added to it.
Felix Fietkau [Wed, 22 Apr 2020 11:10:00 +0000 (13:10 +0200)]
mt76: mt7615: fix tx status rate index calculation
A switch from one rate index to the next only happens when tx count from
the current slot is greater than MT7615_RATE_RETRY, which is 1 has to be
subtracted from count, instead of added to it.
Felix Fietkau [Wed, 22 Apr 2020 15:46:58 +0000 (17:46 +0200)]
mt76: mt7603: never use an 802.11b CF-End rate on 5GHz
Sometimes mt7615_mac_set_timing gets called while the slot time is still
configured to 20. Ensure that in this case it always uses the OFDM CFend
rate.
Felix Fietkau [Wed, 22 Apr 2020 10:25:54 +0000 (12:25 +0200)]
mt76: mt7615: never use an 802.11b CF-End rate on 5GHz
Sometimes mt7615_mac_set_timing gets called while the slot time is still
configured to 20. Ensure that in this case it always uses the OFDM CFend
rate.
Felix Fietkau [Mon, 20 Apr 2020 14:34:16 +0000 (16:34 +0200)]
mt76: fix A-MPDU density handling
The hardware requirements for A-MPDU density are entirely on the tx side,
not the rx side. Because of that, the IE value should stay at 0 and the
minimum value should instead be enforced in WTBL/TXWI
Lorenzo Bianconi [Wed, 22 Apr 2020 08:47:23 +0000 (10:47 +0200)]
mt76: mt7615: rework mt7615_mac_sta_poll for usb code
Since usb code can't access device registers in interrupt context, move
rcu_read_lock/rcu_read_unlock in mt7615_poll_tx routine. Moreover loop
over a local msta list in mt7615_mac_sta_poll since mt7663u driver will
not be able to complete the inner while loop before sta_poll_list list
is refilled by mt7615_mac_add_txs/mt7615_mac_fill_rx
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Mon, 20 Apr 2020 12:39:02 +0000 (14:39 +0200)]
mt76: mt7615: fix max wtbl size for 7663
Current mt7663 offload firmware can support up to 32 wtbl entries
Fixes: f40ac0f3d3c0 ("mt76: mt7615: introduce mt7663e support") Co-developed-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Mon, 20 Apr 2020 11:58:44 +0000 (13:58 +0200)]
mt76: mt7615: parse mcu return code for unified commands
Add return code parsing for the following unified commands:
- MCU_UNI_CMD_DEV_INFO_UPDATE
- MCU_UNI_CMD_BSS_INFO_UPDATE
- MCU_UNI_CMD_STA_REC_UPDATE
Co-developed-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Sean Wang [Fri, 17 Apr 2020 11:13:28 +0000 (13:13 +0200)]
mt76: mt7615: make Kconfig entry obvious for MT7663E
Make Kconfig entry obvious for MT7663E
Signed-off-by: Sean Wang <sean.wang@mediatek.com> Co-developed-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Fri, 17 Apr 2020 11:10:06 +0000 (13:10 +0200)]
mt76: mt7663: introduce 802.11 PS support in sta mode
Enable 802.11 power-save support available in mt7663 firmware
Co-developed-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Fri, 17 Apr 2020 11:10:05 +0000 (13:10 +0200)]
mt76: remove PS_NULLFUNC_STACK capability
remove IEEE80211_HW_PS_NULLFUNC_STACK capability from mt76_phy_init
routine since 802.11 ps is not currently supported by any devices and it
will conflict with mt7663 ps fw support
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Fri, 17 Apr 2020 11:10:04 +0000 (13:10 +0200)]
mt76: mt7615: provide aid info to the mcu
For sta mode mac80211 provides aid in vif->bss_conf.aid.
In order to properly support 802.11 power-save, configure correct aid
to mcu during sta association
Co-developed-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Thu, 16 Apr 2020 14:32:51 +0000 (16:32 +0200)]
mt76: mt7615: introduce mt7663u support
Introduce support for mt7663u 802.11ac 2x2:2 chipset to mt7615 driver.
Main difference respect to pcie code base is the usb code needs to
configure wtbl from non-atomic context
Co-developed-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Thu, 16 Apr 2020 14:32:50 +0000 (16:32 +0200)]
mt76: mt7615: move core shared code in mt7615-common module
Create mt7615-common module in order to collect shared code between usb
and mmio code. Move the following source files in mt7615-common module:
- main.c
- init.c
- mcu.c
- mac.c
- debugfs.c
- eeprom.c
- trace.c
Create the following source files for mmio only source code and move them
in mt7615e module:
- pci_init.c
- dma.c
- pci_mac.c
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Mon, 13 Apr 2020 14:28:48 +0000 (16:28 +0200)]
mt76: mt7615: fix possible deadlock in mt7615_stop
make mac_work per phy instead of per device and fix a possible deadlock
in mt7615_stop since mt7615_mac_work runs holding mt76 mutex
Fixes: fdd2e570764c2 ("mt76: mt7615: add dual-phy support for mac80211 ops") Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Sparse report warnings at mt76_tx_status_unlock() and mt76_tx_status_lock()
warning: context imbalance in mt76_tx_status_lock() - wrong count at exit
warning: context imbalance in mt76_tx_status_unlock() - unexpected unlock
The root cause is the additional __acquire(&dev->status_list.lock)
and __release(&dev->status_list.unlock) called
from inside mt76_tx_status_lock() and mt76_tx_status_unlock().
Remove __acquire(&dev->status_list.lock) annotation
Remove __releases(&dev->status_list.unlock)
Correct &dev->status_list.unlock to &dev->status_list.lock
-unlock not defined in the sk_buff_head struct
Signed-off-by: Jules Irenge <jbi.octave@gmail.com> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Felix Fietkau [Thu, 9 Apr 2020 12:37:50 +0000 (14:37 +0200)]
mt76: mt7615: rework IRQ handling to prepare for MSI support
With MSI interrupts, IRQs must not be enabled from within the IRQ handler,
because that can lead to lost events.
Defer IRQ processing to a tasklet, which is also responsible for enabling
IRQs (to avoid race conditions against the handler)
Co-developed-by: Soul Huang <Soul.Huang@mediatek.com> Acked-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Soul Huang <soul.huang@mediatek.com> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Felix Fietkau [Mon, 6 Apr 2020 12:03:06 +0000 (14:03 +0200)]
mt76: mt76x2: disable merge of OTP ROM data by default
The reference driver does not seem to enable it by default, only under certain
conditions, e.g. when a .bin file is loaded.
Make it opt-in via a device tree property for now, in case it is needed on some
boards.
Felix Fietkau [Mon, 6 Apr 2020 12:01:56 +0000 (14:01 +0200)]
mt76: mt7603: disable merge of OTP ROM data by default
The reference driver does not seem to enable it by default, only under certain
conditions, e.g. when a .bin file is loaded.
Make it opt-in via a device tree property for now, in case it is needed on some
boards.
Felix Fietkau [Mon, 6 Apr 2020 08:19:48 +0000 (10:19 +0200)]
mt76: mt7615: disable merge of OTP ROM data by default
The reference driver does not seem to enable it by default, only under certain
conditions, e.g. when a .bin file is loaded.
Make it opt-in via a device tree property for now, in case it is needed on some
boards.
Lorenzo Bianconi [Fri, 3 Apr 2020 19:01:56 +0000 (21:01 +0200)]
mt76: mt7615: do not always reset the dfs state setting the channel
mac80211/hostapd runs mt7615_set_channel with the same channel
parameters sending multiple rdd commands overwriting the previous ones.
This behaviour is causing tpt issues on dfs channels.
Fix the issue checking new channel freq/width with the running one.
Fixes: 5dabdf71e94e ("mt76: mt7615: add multiple wiphy support to the dfs support code") Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Thu, 2 Apr 2020 18:18:55 +0000 (20:18 +0200)]
mt76: mt7615: introduce mt7615_wtbl_desc data structure
Generalize mt7615_rate_desc introducing mt7615_wtbl_desc and
mt7615_key_desc data structures in order to configure the hw wtbl
in a non-atomic context for usb devices
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Thu, 2 Apr 2020 18:18:54 +0000 (20:18 +0200)]
mt76: mt7615: rework wtbl key configuration
Remove key dependency from mt7615_mac_wtbl_update_key and export
mt7615_mac_wtbl_update_key, mt7615_mac_wtbl_update_pk and
mt7615_mac_wtbl_update_cipher in order to reuse them in usb code.
Move mt7615_mac_get_cipher in mac.h
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Move register configuration out of mt7615_mac_set_rates since usb
driver can't access device register in interrupt context. Introduce
mt7615_mac_update_rate_desc routine to report rate info to
mt7615_mac_set_rates
Co-developed-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Thu, 2 Apr 2020 18:18:49 +0000 (20:18 +0200)]
mt76: mt7615: introduce mt7663u support to mt7615_write_txwi
Extend mt7615_write_txwi routine to support usb txwi configuration
Co-developed-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Thu, 2 Apr 2020 18:18:48 +0000 (20:18 +0200)]
mt76: add headroom and tailroom to mt76_mcu_ops data structure
Introduce headroom and tailroom to mt76_mcu_ops data structure in order
to unify the routine used for mcu message allocation. This is a
preliminary patch to add mt7663u support
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Remove mt76_wr(dev, MT_CSR(0x010), 0x8208) that would cause
MT_PCIE_IRQ_ENABLE to be disabled; MT_PCIE_IRQ_ENABLE should always keep
on enabled when the driver is running.
0x44064 is a not existing address
Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Thu, 2 Apr 2020 13:06:41 +0000 (15:06 +0200)]
mt76: mt7615: introduce rlm tlv in bss_info mcu command
Introduce rlm tlv header in bss_info mcu command in order to
inform the mcu about operating channel. Rlm header is necessary only if
the mcu is running low power functionalities (e.g offloaded scan)
Co-developed-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Sean Wang [Thu, 2 Apr 2020 13:06:40 +0000 (15:06 +0200)]
mt76: mt7615: introduce BSS absence event
Introduce BSS absence event that is reported when the fw
is leaving or entering current operational channel.
Signed-off-by: Sean Wang <sean.wang@mediatek.com> Co-developed-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Thu, 2 Apr 2020 13:06:39 +0000 (15:06 +0200)]
mt76: mt7615: introduce scheduled scan support
Introduce scheduled scan support for mt7663e devices
Co-developed-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Thu, 2 Apr 2020 13:06:38 +0000 (15:06 +0200)]
mt76: mt7615: introduce hw scan support
Introduce hw scan support to mt7663e driver
Co-developed-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Sean Wang [Thu, 2 Apr 2020 13:06:37 +0000 (15:06 +0200)]
mt76: mt7663: keep Rx filters as the default
Keep Rx filters default value if the firmware supports offload and
low power features.
Signed-off-by: Sean Wang <sean.wang@mediatek.com> Co-developed-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Introduce mt7615_mcu_set_channel_domain routines in order to instruct
the mcu about supported band/channels. This is a preliminary patch to
add hw scan support to mt7663e driver
Co-developed-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Thu, 2 Apr 2020 13:06:34 +0000 (15:06 +0200)]
mt76: mt7663: disable RDD commands
Disable dfs RDD mcu commands for mt7663 driver since they are not
currently supported by the 7663 firmware
Co-developed-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Lorenzo Bianconi [Thu, 2 Apr 2020 13:06:32 +0000 (15:06 +0200)]
mt76: mt7663: enable nf estimation
Enable Noise floor estimation for mt7663 driver
Co-developed-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org> Signed-off-by: Felix Fietkau <nbd@nbd.name>
Felix Fietkau [Sun, 16 Feb 2020 15:08:58 +0000 (16:08 +0100)]
mt76: mt76x02: fix handling MCU timeouts during hw restart
If a MCU timeout occurs before a hw restart completes, another hw restart
is scheduled, and the station state gets corrupted.
To speed up dealing with that, do not issue any MCU commands after the first
timeout, and defer handling timeouts until the reset has completed.
Also ignore errors in MCU commands during start/config to avoid making user
space fail on this condition. If it happens, another restart is scheduled
quickly, and that usually recovers the hardware properly.
Gustavo A. R. Silva [Thu, 7 May 2020 19:26:47 +0000 (14:26 -0500)]
rndis_wlan: Replace zero-length array with flexible-array
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
sizeof(flexible-array-member) triggers a warning because flexible array
members have incomplete type[1]. There are some instances of code in
which the sizeof operator is being incorrectly/erroneously applied to
zero-length arrays and the result is zero. Such instances may be hiding
some bugs. So, this work (flexible-array member conversions) will also
help to get completely rid of those sorts of issues.
Gustavo A. R. Silva [Thu, 7 May 2020 19:19:26 +0000 (14:19 -0500)]
qtnfmac: Replace zero-length array with flexible-array
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
sizeof(flexible-array-member) triggers a warning because flexible array
members have incomplete type[1]. There are some instances of code in
which the sizeof operator is being incorrectly/erroneously applied to
zero-length arrays and the result is zero. Such instances may be hiding
some bugs. So, this work (flexible-array member conversions) will also
help to get completely rid of those sorts of issues.
Signed-off-by: Gustavo A. R. Silva <gustavoars@kernel.org> Reviewed-by: Sergey Matyukevich <sergey.matyukevich.os@quantenna.com> Signed-off-by: Kalle Valo <kvalo@codeaurora.org> Link: https://lore.kernel.org/r/20200507191926.GA15970@embeddedor
Gustavo A. R. Silva [Thu, 7 May 2020 19:02:10 +0000 (14:02 -0500)]
prism54: Replace zero-length array with flexible-array
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
sizeof(flexible-array-member) triggers a warning because flexible array
members have incomplete type[1]. There are some instances of code in
which the sizeof operator is being incorrectly/erroneously applied to
zero-length arrays and the result is zero. Such instances may be hiding
some bugs. So, this work (flexible-array member conversions) will also
help to get completely rid of those sorts of issues.
Gustavo A. R. Silva [Thu, 7 May 2020 18:59:14 +0000 (13:59 -0500)]
mwl8k: Replace zero-length array with flexible-array
The current codebase makes use of the zero-length array language
extension to the C90 standard, but the preferred mechanism to declare
variable-length types such as these ones is a flexible array member[1][2],
introduced in C99:
struct foo {
int stuff;
struct boo array[];
};
By making use of the mechanism above, we will get a compiler warning
in case the flexible array does not occur last in the structure, which
will help us prevent some kind of undefined behavior bugs from being
inadvertently introduced[3] to the codebase from now on.
Also, notice that, dynamic memory allocations won't be affected by
this change:
"Flexible array members have incomplete type, and so the sizeof operator
may not be applied. As a quirk of the original implementation of
zero-length arrays, sizeof evaluates to zero."[1]
sizeof(flexible-array-member) triggers a warning because flexible array
members have incomplete type[1]. There are some instances of code in
which the sizeof operator is being incorrectly/erroneously applied to
zero-length arrays and the result is zero. Such instances may be hiding
some bugs. So, this work (flexible-array member conversions) will also
help to get completely rid of those sorts of issues.
projects / linux.git / log