diff options
| author | Po-Hao Huang <phhuang@realtek.com> | 2021-04-15 16:47:03 +0800 |
|---|---|---|
| committer | Kalle Valo <kvalo@codeaurora.org> | 2021-04-21 12:37:50 +0300 |
| commit | 7bd3760c71f7a18485d2c10ea0887e1d41519f4e (patch) | |
| tree | a6a1e9d7a062704bfbec60d1442ca788029562a0 /drivers/net/wireless/ti | |
| parent | a548909d7ad7e334c6c923a71f0b694d60980232 (diff) | |
rtw88: refine napi deinit flow
We used to stop napi before disabling irqs. And it turns out
to cause some problem when we try to stop device while interrupt arrives.
To safely stop pci, we do three steps:
1. disable interrupt
2. synchronize_irq
3. stop_napi
Since step 2 and 3 may not finish as expected when interrupt is enabled,
use rtwpci->running to decide whether interrupt should be re-enabled at
the time.
Fixes: 9e2fd29864c5 ("rtw88: add napi support")
Signed-off-by: Po-Hao Huang <phhuang@realtek.com>
Signed-off-by: Ping-Ke Shih <pkshih@realtek.com>
Signed-off-by: Kalle Valo <kvalo@codeaurora.org>
Link: https://lore.kernel.org/r/20210415084703.27255-4-pkshih@realtek.com
Diffstat (limited to 'drivers/net/wireless/ti')
0 files changed, 0 insertions, 0 deletions