On Sat, Jan 17, 2026 at 10:10:09AM +0800, Shawn Lin wrote:

在 2026/01/16 星期五 23:37, Mark Brown 写道:

quoted

TBH it looked to me more like there'd been some deferral of more complex
work at some point but that didn't make it into the final code.  In

Ah, indeed，rockchip_spi_isr() directly performs R/W pio in hardirq
context which doesn't seem advisable, so probably some better
improvement is needed in the furture. Anyway, fix the warning right
now as $subject patch looks sensible:

Generally the best pattern with this stuff is to do PIO in process
context - it looks like the driver already has a copybreak to use DMA
for larger transfers on systems that have DMA support so this should
only be happening for very short transfers in which case the context
thrashing from going to interrupt context is probably hurting
performance anyway.