From 5c56591b57bd4540cdab4cd02d7c07f8544ca967 Mon Sep 17 00:00:00 2001 From: Rich Ercolani <214141+rincebrain@users.noreply.github.com> Date: Mon, 20 Sep 2021 13:30:50 -0400 Subject: [PATCH] Handle partial reads in zfs_read Currently, dmu_read_uio_dnode can read 64K of a requested 1M in one loop, get EFAULT back from zfs_uiomove() (because the iovec only holds 64k), and return EFAULT, which turns into EAGAIN on the way out. EAGAIN gets interpreted as "I didn't read anything", the caller tries again without consuming the 64k we already read, and we're stuck. This apparently works on newer kernels because the caller which breaks on older Linux kernels by happily passing along a 1M read request and a 64k iovec just requests 64k at a time. With this, we now won't return EFAULT if we got a partial read. Reviewed-by: Brian Behlendorf Signed-off-by: Rich Ercolani Closes #12370 Closes #12509 Closes #12516 --- module/zfs/zfs_vnops.c | 8 ++++++++ 1 file changed, 8 insertions(+) diff --git a/module/zfs/zfs_vnops.c b/module/zfs/zfs_vnops.c index 1b32d38767..b9e93791d5 100644 --- a/module/zfs/zfs_vnops.c +++ b/module/zfs/zfs_vnops.c @@ -261,6 +261,7 @@ zfs_read(struct znode *zp, zfs_uio_t *uio, int ioflag, cred_t *cr) } ASSERT(zfs_uio_offset(uio) < zp->z_size); + ssize_t start_offset = zfs_uio_offset(uio); ssize_t n = MIN(zfs_uio_resid(uio), zp->z_size - zfs_uio_offset(uio)); ssize_t start_resid = n; @@ -283,6 +284,13 @@ zfs_read(struct znode *zp, zfs_uio_t *uio, int ioflag, cred_t *cr) /* convert checksum errors into IO errors */ if (error == ECKSUM) error = SET_ERROR(EIO); + /* + * if we actually read some bytes, bubbling EFAULT + * up to become EAGAIN isn't what we want here. + */ + if (error == EFAULT && + (zfs_uio_offset(uio) - start_offset) != 0) + error = 0; break; }