On 9/7/25 7:35 AM, Jeff Layton wrote:
> On Sat, 2025-09-06 at 17:25 -0400, Chuck Lever wrote:
>> From: Mike Snitzer <snitzer@xxxxxxxxxx>
>>
>> Add 'io_cache_read' to NFSD's debugfs interface so that any data
>> read by NFSD will either be:
>> - cached using page cache (NFSD_IO_BUFFERED=0)
>> - cached but removed from the page cache upon completion
>> (NFSD_IO_DONTCACHE=1).
>>
>> io_cache_read may be set by writing to:
>> /sys/kernel/debug/nfsd/io_cache_read
>>
>> Add 'io_cache_write' to NFSD's debugfs interface so that any data
>> written by NFSD will either be:
>> - cached using page cache (NFSD_IO_BUFFERED=0)
>> - cached but removed from the page cache upon completion
>> (NFSD_IO_DONTCACHE=1).
>>
>> io_cache_write may be set by writing to:
>> /sys/kernel/debug/nfsd/io_cache_write
>>
>> The default value for both settings is NFSD_IO_BUFFERED, which is
>> NFSD's existing behavior for both read and write. Changes to these
>> settings take immediate effect for all exports and NFS versions.
>>
>> Currently only xfs and ext4 implement RWF_DONTCACHE. For file
>> systems that do not implement RWF_DONTCACHE, NFSD use only buffered
>> I/O when the io_cache setting is NFSD_IO_DONTCACHE.
>>
>> Signed-off-by: Mike Snitzer <snitzer@xxxxxxxxxx>
>> Signed-off-by: Chuck Lever <chuck.lever@xxxxxxxxxx>
>> ---
>> fs/nfsd/debugfs.c | 93 +++++++++++++++++++++++++++++++++++++++++++++++
>> fs/nfsd/nfsd.h | 9 +++++
>> fs/nfsd/vfs.c | 21 +++++++++++
>> 3 files changed, 123 insertions(+)
>>
>> Changes since v1:
>> - Corrected patch author
>> - Break back to NFSD_IO_BUFFERED when exported file system does not
>> support RWF_DONTCACHE
>> - Smoke tested NFSD_IO_DONTCACHE with NFSv3,v4.0,v4.1 on xfs and
>> tmpfs
>>
>> diff --git a/fs/nfsd/debugfs.c b/fs/nfsd/debugfs.c
>> index 84b0c8b559dc..2b1bb716b608 100644
>> --- a/fs/nfsd/debugfs.c
>> +++ b/fs/nfsd/debugfs.c
>> @@ -27,11 +27,98 @@ static int nfsd_dsr_get(void *data, u64 *val)
>> static int nfsd_dsr_set(void *data, u64 val)
>> {
>> nfsd_disable_splice_read = (val > 0) ? true : false;
>> + if (!nfsd_disable_splice_read) {
>> + /*
>> + * Must use buffered I/O if splice_read is enabled.
>> + */
>> + nfsd_io_cache_read = NFSD_IO_BUFFERED;
>> + }
>> return 0;
>> }
>>
>> DEFINE_DEBUGFS_ATTRIBUTE(nfsd_dsr_fops, nfsd_dsr_get, nfsd_dsr_set, "%llu\n");
>>
>> +/*
>> + * /sys/kernel/debug/nfsd/io_cache_read
>> + *
>> + * Contents:
>> + * %0: NFS READ will use buffered IO
>> + * %1: NFS READ will use dontcache (buffered IO w/ dropbehind)
>> + *
>> + * This setting takes immediate effect for all NFS versions,
>> + * all exports, and in all NFSD net namespaces.
>> + */
>> +
>> +static int nfsd_io_cache_read_get(void *data, u64 *val)
>> +{
>> + *val = nfsd_io_cache_read;
>> + return 0;
>> +}
>> +
>> +static int nfsd_io_cache_read_set(void *data, u64 val)
>> +{
>> + int ret = 0;
>> +
>> + switch (val) {
>> + case NFSD_IO_BUFFERED:
>> + nfsd_io_cache_read = NFSD_IO_BUFFERED;
>> + break;
>> + case NFSD_IO_DONTCACHE:
>> + /*
>> + * Must disable splice_read when enabling
>> + * NFSD_IO_DONTCACHE.
>> + */
>> + nfsd_disable_splice_read = true;
>> + nfsd_io_cache_read = val;
>> + break;
>> + default:
>> + ret = -EINVAL;
>> + break;
>> + }
>> +
>> + return ret;
>> +}
>> +
>> +DEFINE_DEBUGFS_ATTRIBUTE(nfsd_io_cache_read_fops, nfsd_io_cache_read_get,
>> + nfsd_io_cache_read_set, "%llu\n");
>> +
>> +/*
>> + * /sys/kernel/debug/nfsd/io_cache_write
>> + *
>> + * Contents:
>> + * %0: NFS WRITE will use buffered IO
>> + * %1: NFS WRITE will use dontcache (buffered IO w/ dropbehind)
>> + *
>> + * This setting takes immediate effect for all NFS versions,
>> + * all exports, and in all NFSD net namespaces.
>> + */
>> +
>> +static int nfsd_io_cache_write_get(void *data, u64 *val)
>> +{
>> + *val = nfsd_io_cache_write;
>> + return 0;
>> +}
>> +
>> +static int nfsd_io_cache_write_set(void *data, u64 val)
>> +{
>> + int ret = 0;
>> +
>> + switch (val) {
>> + case NFSD_IO_BUFFERED:
>> + case NFSD_IO_DONTCACHE:
>> + nfsd_io_cache_write = val;
>> + break;
>> + default:
>> + ret = -EINVAL;
>> + break;
>> + }
>> +
>> + return ret;
>> +}
>> +
>> +DEFINE_DEBUGFS_ATTRIBUTE(nfsd_io_cache_write_fops, nfsd_io_cache_write_get,
>> + nfsd_io_cache_write_set, "%llu\n");
>> +
>> void nfsd_debugfs_exit(void)
>> {
>> debugfs_remove_recursive(nfsd_top_dir);
>> @@ -44,4 +131,10 @@ void nfsd_debugfs_init(void)
>>
>> debugfs_create_file("disable-splice-read", S_IWUSR | S_IRUGO,
>> nfsd_top_dir, NULL, &nfsd_dsr_fops);
>> +
>> + debugfs_create_file("io_cache_read", 0644, nfsd_top_dir, NULL,
>> + &nfsd_io_cache_read_fops);
>> +
>> + debugfs_create_file("io_cache_write", 0644, nfsd_top_dir, NULL,
>> + &nfsd_io_cache_write_fops);
>> }
>> diff --git a/fs/nfsd/nfsd.h b/fs/nfsd/nfsd.h
>> index 1cd0bed57bc2..809729d41e08 100644
>> --- a/fs/nfsd/nfsd.h
>> +++ b/fs/nfsd/nfsd.h
>> @@ -153,6 +153,15 @@ static inline void nfsd_debugfs_exit(void) {}
>>
>> extern bool nfsd_disable_splice_read __read_mostly;
>>
>> +enum {
>> + /* Any new NFSD_IO enum value must be added at the end */
>> + NFSD_IO_BUFFERED,
>> + NFSD_IO_DONTCACHE,
>> +};
>> +
>> +extern u64 nfsd_io_cache_read __read_mostly;
>> +extern u64 nfsd_io_cache_write __read_mostly;
>> +
>> extern int nfsd_max_blksize;
>>
>> static inline int nfsd_v4client(struct svc_rqst *rq)
>> diff --git a/fs/nfsd/vfs.c b/fs/nfsd/vfs.c
>> index 3cd3b9e069f4..714777c221ed 100644
>> --- a/fs/nfsd/vfs.c
>> +++ b/fs/nfsd/vfs.c
>> @@ -49,6 +49,8 @@
>> #define NFSDDBG_FACILITY NFSDDBG_FILEOP
>>
>> bool nfsd_disable_splice_read __read_mostly;
>> +u64 nfsd_io_cache_read __read_mostly = NFSD_IO_BUFFERED;
>> +u64 nfsd_io_cache_write __read_mostly = NFSD_IO_BUFFERED;
>>
>> /**
>> * nfserrno - Map Linux errnos to NFS errnos
>> @@ -1099,6 +1101,16 @@ __be32 nfsd_iter_read(struct svc_rqst *rqstp, struct svc_fh *fhp,
>> size_t len;
>>
>> init_sync_kiocb(&kiocb, file);
>> +
>> + switch (nfsd_io_cache_read) {
>> + case NFSD_IO_BUFFERED:
>> + break;
>> + case NFSD_IO_DONTCACHE:
>> + if (file->f_op->fop_flags & FOP_DONTCACHE)
>> + kiocb.ki_flags = IOCB_DONTCACHE;
>> + break;
>> + }
>> +
>> kiocb.ki_pos = offset;
>>
>> v = 0;
>> @@ -1224,6 +1236,15 @@ nfsd_vfs_write(struct svc_rqst *rqstp, struct svc_fh *fhp,
>> since = READ_ONCE(file->f_wb_err);
>> if (verf)
>> nfsd_copy_write_verifier(verf, nn);
>> +
>> + switch (nfsd_io_cache_write) {
>> + case NFSD_IO_BUFFERED:
>> + break;
>> + case NFSD_IO_DONTCACHE:
>> + if (file->f_op->fop_flags & FOP_DONTCACHE)
>> + kiocb.ki_flags |= IOCB_DONTCACHE;
>> + break;
>> + }
>> host_err = vfs_iocb_iter_write(file, &kiocb, &iter);
>> if (host_err < 0) {
>> commit_reset_write_verifier(nn, rqstp, host_err);
>
> Could we get this into v6.18?
Barring any new issues that come up, I'm aiming this one for v6.18.
I have rewritten the patch that does direct reads, but it's not quite
ready yet for human consumption. I can't promise it will be ready for
v6.18.
> Reviewed-by: Jeff Layton <jlayton@xxxxxxxxxx>
Thanks!
--
Chuck Lever
