On Fri, May 09, 2025 at 01:01:21PM +0200, Corentin Labbe wrote: > > Got http://kernel.montjoie.ovh/478064.log OK it seems to go further, but that might just because the first few tests get done with no concurrency. I see some bugs in the driver with regards to zero-length updates, but I'm at a loss as to why that would only make a difference when run concurrently. Please add this patch on top of the previous patch. It disables chaining altogether between requests. Thanks, -- Email: Herbert Xu <herbert@xxxxxxxxxxxxxxxxxxx> Home Page: http://gondor.apana.org.au/~herbert/ PGP Key: http://gondor.apana.org.au/~herbert/pubkey.txt -- diff --git a/drivers/crypto/marvell/cesa/tdma.c b/drivers/crypto/marvell/cesa/tdma.c index 9b5fd957dde2..f4fe2a881f1e 100644 --- a/drivers/crypto/marvell/cesa/tdma.c +++ b/drivers/crypto/marvell/cesa/tdma.c @@ -123,7 +123,7 @@ void mv_cesa_tdma_chain(struct mv_cesa_engine *engine, * Break the DMA chain if the CESA_TDMA_BREAK_CHAIN is set on * the last element of the current chain. */ - if (last->flags & CESA_TDMA_BREAK_CHAIN) { + if (1 || (last->flags & CESA_TDMA_BREAK_CHAIN)) { engine->chain_sw.first = NULL; engine->chain_sw.last = NULL; }
