Remove following struct no longer used:
- dev_info
- bss_info
- bss_info_tag_handler
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Remove function pointers in mt7615_mcu_set_bss_info and run function
directly. Moreover remove __mt7615_mcu_set_bss_info since it is run just
by mt7615_mcu_set_bss_info and remove duplicated istructions
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Unify mt7615_mcu_set_dev_info and __mt7615_mcu_set_dev_info since the
latter is run just by mt7615_mcu_set_dev_info
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Remove skb_ret parameter from mt7615_mcu_msg_send signature since it is
actually used just by mt7615_mcu_patch_sem_ctrl. This is a prelimanry
patch to use mt76 common mcu API
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Remove dest parameter from mt7615_mcu_msg_send/__mt7615_mcu_msg_send
routine signature since it can is always set to MCU_S2D_H2N
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Remove query parameter from mt7615_mcu_msg_send routine signature since
it can be obtained from cmd value
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Remove unused vif parameter in mt7615_mcu_del_wtbl signature
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Use proper mcu message size in mt7615_mcu_add_wtbl and do not
allocate a huge buffer. Moreover use stack memory instead of heap one
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Use proper mcu message size in mt7615_mcu_add_wtbl_bmc and do not
allocate a huge buffer. Moreover use stack memory instead of heap one
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Remove bss_info_convert_vif_type routine since it is run just in
mt7615_mcu_set_bss_info and the switch over vif->type is already there.
Simplify mt7615_mcu_set_bss_info routine
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Do not loop over cmd payload in __mt7615_mcu_set_sta_rec since it is
already done in before running __mt7615_mcu_set_sta_rec (e.g.
mt7615_mcu_set_sta_rec)
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Do not loop over cmd payload in __mt7615_mcu_set_wtbl since it is
already done in before running __mt7615_mcu_set_wtbl (e.g.
mt7615_mcu_set_wtbl_key)
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Fix sparse warning:
drivers/net/wireless/mediatek/mt76/mt7615/pci.c:37:13:
warning: symbol 'mt7615_irq_handler' was not declared. Should it be static?
Signed-off-by: YueHaibing <yuehaibing@huawei.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
We dereference "skb" when we assign:
req.pkt_len = cpu_to_le16(MT_TXD_SIZE + skb->len);
^^^^^^^^
So this patch just moves the dev_kfree_skb() down a bit to avoid the
use after free.
Fixes: 04b8e65922 ("mt76: add mac80211 driver for MT7615 PCIe-based chipsets")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Acked-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
The problem is that "mvif->omac_idx" is a u8 so it can't be negative
and the error handling won't work. The get_omac_idx() function returns
-1 on error.
Fixes: 04b8e65922 ("mt76: add mac80211 driver for MT7615 PCIe-based chipsets")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Move netif_napi_del in mt76_dma_cleanup routine since it is done
by all drivers
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
This allows tx scheduling and tx cleanup to run concurrently
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
This allows tx scheduling and tx cleanup to run concurrently
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Move tx_napi in mt76_dev data structure in order to implement
concurrency between tx scheduling and tx cleanup in mt7603 and mt7615
drivers
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Remove useless return statment in mt76x02_resync_beacon_timer routine
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Signed-off-by: Felix Fietkau <nbd@nbd.name>
The current implementation of the event queue both
wastes space using a doubly linked list and isn't super
obvious in how it behaves. This converts the queue to an
actual circular buffer. The size of the queue is a
tunable module parameter. This also fixes a few other things:
- A memory leak that occurred when the ACPI device was
removed, but the events were not freed from the queue.
- Now kfree() the oldest event from outside all locks.
- Add newline to logging messages.
- Add helper macros to calculate size of events.
- Remove unneeded lock around a check for dev_data->exist
in hangup_device().
- Remove an unneeded null event pointer check in enqueue_events().
- Correct some comments.
Signed-off-by: Nick Crews <ncrews@chromium.org>
Signed-off-by: Enric Balletbo i Serra <enric.balletbo@collabora.com>
What is the point in surrounding the whole of declarations with
ifdef like this?
#ifdef CONFIG_FOO
int foo(void);
#endif
If CONFIG_FOO is not defined, all callers of foo() will fail
with implicit declaration errors since the top Makefile adds
-Werror-implicit-function-declaration to KBUILD_CFLAGS.
This breaks the build earlier when you are doing something wrong.
That's it.
Anyway, it will fail to link since the definition of foo() is not
compiled.
In summary, these ifdef are unneeded.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Otherwise, selecting it without MODULES leads to build failures.
Fixes: 58557e486f ("arm64: Allow user selection of ARM64_MODULE_PLTS")
Signed-off-by: Catalin Marinas <catalin.marinas@arm.com>
Currently pt_regs on x86_32 has an oddity in that kernel regs
(!user_mode(regs)) are short two entries (esp/ss). This means that any
code trying to use them (typically: regs->sp) needs to jump through
some unfortunate hoops.
Change the entry code to fix this up and create a full pt_regs frame.
This then simplifies various trampolines in ftrace and kprobes, the
stack unwinder, ptrace, kdump and kgdb.
Much thanks to Josh for help with the cleanups!
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Josh Poimboeuf <jpoimboe@redhat.com>
Acked-by: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The kprobe trampolines have a FRAME_POINTER annotation that makes no
sense. It marks the frame in the middle of pt_regs, at the place of
saving BP.
Change it to mark the pt_regs frame as per the ENCODE_FRAME_POINTER
from the respective entry_*.S.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Josh Poimboeuf <jpoimboe@redhat.com>
Acked-by: Masami Hiramatsu <mhiramat@kernel.org>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
In preparation for wider use, move the ENCODE_FRAME_POINTER macros to
a common header and provide inline asm versions.
These macros are used to encode a pt_regs frame for the unwinder; see
unwind_frame.c:decode_frame_pointer().
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
The code flow around the return from interrupt preemption point seems
needlessly complicated.
There is only one site jumping to resume_kernel, and none (outside of
resume_kernel) jumping to restore_all_kernel. Inline resume_kernel
in restore_all_kernel and avoid the CONFIG_PREEMPT dependent label.
Signed-off-by: Peter Zijlstra (Intel) <peterz@infradead.org>
Reviewed-by: Josh Poimboeuf <jpoimboe@redhat.com>
Cc: Linus Torvalds <torvalds@linux-foundation.org>
Cc: Peter Zijlstra <peterz@infradead.org>
Cc: Thomas Gleixner <tglx@linutronix.de>
Signed-off-by: Ingo Molnar <mingo@kernel.org>
Without 'set -e', shell scripts continue running even after any
error occurs. The missed 'set -e' is a typical bug in shell scripting.
For example, when a disk space shortage occurs while this script is
running, it actually ends up with generating a truncated capflags.c.
Yet, mkcapflags.sh continues running and exits with 0. So, the build
system assumes it has succeeded.
It will not be re-generated in the next invocation of Make since its
timestamp is newer than that of any of the source files.
Add 'set -e' so that any error in this script is caught and propagated
to the build system.
Since 9c2af1c737 ("kbuild: add .DELETE_ON_ERROR special target"),
make automatically deletes the target on any failure. So, the broken
capflags.c will be deleted automatically.
Signed-off-by: Masahiro Yamada <yamada.masahiro@socionext.com>
Signed-off-by: Thomas Gleixner <tglx@linutronix.de>
Cc: "H. Peter Anvin" <hpa@zytor.com>
Cc: Borislav Petkov <bp@alien8.de>
Link: https://lkml.kernel.org/r/20190625072622.17679-1-yamada.masahiro@socionext.com
Check if we need to allocate uncached memory for a device given the
allocation flags. Switch over the uncached segment check to this helper
to deal with architectures that do not support the dma_cache_sync
operation and thus should not returned cacheable memory for
DMA_ATTR_NON_CONSISTENT allocations.
Signed-off-by: Christoph Hellwig <hch@lst.de>
The openrisc DMA code supports DMA_ATTR_NON_CONSISTENT allocations, but
does not provide a cache_sync operation. This means any user of it
will never be able to actually transfer cache ownership and thus cause
coherency bugs.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Acked-by: Stafford Horne <shorne@gmail.com>
The arc DMA code supports DMA_ATTR_NON_CONSISTENT allocations, but does
not provide a cache_sync operation. This means any user of it will
never be able to actually transfer cache ownership and thus cause
coherency bugs.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Evgeniy Paltsev <paltsev@synopsys.com>
Tested-by: Evgeniy Paltsev <paltsev@synopsys.com>
The arm-nommu DMA code supports DMA_ATTR_NON_CONSISTENT allocations, but
does not provide a cache_sync operation. This means any user of it
will never be able to actually transfer cache ownership and thus cause
coherency bugs.
Signed-off-by: Christoph Hellwig <hch@lst.de>
Reviewed-by: Vladimir Murzin <vladimir.murzin@arm.com>
Since the Linux 5.1 merge window we allow drivers to just set the
largest DMA mask they support instead of falling back to smaller ones.
But I forgot to remove a check that prohibits this behavior in the
arm DMA code, as it is rather hidden. There is not reason for this check
as the code will do the right thing for a "too large" DMA mask, so
just remove it.
Fixes: 9eb9e96e97 ("Documentation/DMA-API-HOWTO: update dma_mask sections")
Signed-off-by: Christoph Hellwig <hch@lst.de>
The dma masks in struct device are always 64-bits wide. But for builds
using a 32-bit dma_addr_t we need to ensure we don't store an
unsupportable value. Before Linux 5.0 this was handled at least by
the ARM dma mapping code by never allowing to set a larger dma_mask,
but these days we allow the driver to just set the largest supported
value and never fall back to a smaller one. Ensure this always works
by truncating the value.
Fixes: 9eb9e96e97 ("Documentation/DMA-API-HOWTO: update dma_mask sections")
Signed-off-by: Christoph Hellwig <hch@lst.de>
