Re: [PATCHv3 1/8] block: check for valid bio while splitting
From: Christoph Hellwig <hch@lst.de>
Date: 2025-08-25 07:35:50
Also in:
linux-fsdevel, linux-xfs, lkml
On Tue, Aug 19, 2025 at 09:49:15AM -0700, Keith Busch wrote:
quoted hunk ↗ jump to hunk
/* * If the queue doesn't support SG gaps and adding this * offset would create a gap, disallow it.@@ -339,8 +343,16 @@ int bio_split_rw_at(struct bio *bio, const struct queue_limits *lim, * Individual bvecs might not be logical block aligned. Round down the * split size so that each bio is properly block size aligned, even if * we do not use the full hardware limits. + * + * Misuse may submit a bio that can't be split into a valid io. There + * may either be too many discontiguous vectors for the max segments + * limit, or contain virtual boundary gaps without having a valid block + * sized split. Catch that condition by checking for a zero byte + * result. */ bytes = ALIGN_DOWN(bytes, bio_split_alignment(bio, lim)); + if (!bytes)
If this is just misuse it could be a WARN_ON_ONCE. But I think we can also trigger this when validating passthrough commands that need to be built to hardware limits. So maybe don't speak about misuse here? Otherwise looks good: Reviewed-by: Christoph Hellwig <hch@lst.de>