Re: [PATCH] spi: axi-spi-engine: fix stale SYNC IRQ pending

[Jonathan Santos]

On 09/03, Andy Shevchenko wrote:
> On Thu, Sep 03, 2026 at 10:49:22AM -0300, Jonathan Santos wrote:
> > spi_engine_setup() sends a SYNC(1) command and polls SYNC_ID to confirm
> > it was parsed by the FPGA, but never clears the corresponding interrupt
> > pending bit (INT_PENDING[SYNC]). When the first real SPI transfer starts
> > and INT_SYNC is enabled, that stale pending bit fires immediately, causing
> > the IRQ handler to see the leftover SYNC_ID from setup, match it against
> > the current transfer's ID, and prematurely signal completion before the
> > hardware finishes.
> > 
> > This race manifests at low SPI clock frequencies (~2-3 MHz), where the
> > FPGA takes long enough to execute the transfer that handler is parsed
> > before it finishes. At higher SCLK rates the transfer completes fast
> > enough that the issue is masked.
> > 
> > Fix this by clearing INT_PENDING[SYNC] after the polled SYNC, ensuring no
> > stale interrupt is left pending.
> > 
> > Reported-by: Dennis Heinzel <dennis.heinzel@irs.systems>
> > Link: https://ez.analog.com/linux-software-drivers/f/q-a/604145/axi-spi-engine-stale-sync-pending-can-complete-first-transfer-early-at-low-spi-clock-2-3-mhz
> 
> We have a Closes tag.
> 
> > Signed-off-by: Jonathan Santos <Jonathan.Santos@xxxxxxxxxx>
> 
> ...
> 
> > +	ret = readl_relaxed_poll_timeout(spi_engine->base + SPI_ENGINE_REG_SYNC_ID,
> > +					 reg, reg == 1, 1, 1000);
> 
> While at it I would replace 1000 with 1 * USEC_PER_MSEC
> 
> > +	/* Clear the stale SYNC pending bit so it doesn't fire when the IRQ is later enabled */
> > +	writel_relaxed(SPI_ENGINE_INT_SYNC, spi_engine->base + SPI_ENGINE_REG_INT_PENDING);
> 
> In both cases? Error (timeout) and not?
> 

The error (timeout) indicates the SYNC command was not parsed within the
deadline, but it can be executed at any time. We consider the timeout big
enough, so this is unlikely to happen. But in any case, the cpu command to
clear INT_PENDING is harmeless and can still clear the interrupt if the 
SYNC is done parsing until right before this command is executed.

> > +	return ret;
> 
> -- 
> With Best Regards,
> Andy Shevchenko
>