From: Aurelien Aptel aaptel@suse.com
[ Upstream commit 88fd98a2306755b965e4f4567f84e73db3b6738c ]
When doing a large read or write workload we only very gradually increase the number of credits which can cause problems with parallelizing large i/o (I/O ramps up more slowly than it should for large read/write workloads) especially with multichannel when the number of credits on the secondary channels starts out low (e.g. less than about 130) or when recovering after server throttled back the number of credit.
Signed-off-by: Aurelien Aptel aaptel@suse.com Reviewed-by: Shyam Prasad N sprasad@microsoft.com Signed-off-by: Steve French stfrench@microsoft.com Signed-off-by: Sasha Levin sashal@kernel.org --- fs/cifs/smb2pdu.c | 6 ++---- 1 file changed, 2 insertions(+), 4 deletions(-)
diff --git a/fs/cifs/smb2pdu.c b/fs/cifs/smb2pdu.c index c6f8bc6729aa..d1d550647cd6 100644 --- a/fs/cifs/smb2pdu.c +++ b/fs/cifs/smb2pdu.c @@ -4032,8 +4032,7 @@ smb2_async_readv(struct cifs_readdata *rdata) if (rdata->credits.value > 0) { shdr->CreditCharge = cpu_to_le16(DIV_ROUND_UP(rdata->bytes, SMB2_MAX_BUFFER_SIZE)); - shdr->CreditRequest = - cpu_to_le16(le16_to_cpu(shdr->CreditCharge) + 1); + shdr->CreditRequest = cpu_to_le16(le16_to_cpu(shdr->CreditCharge) + 8);
rc = adjust_credits(server, &rdata->credits, rdata->bytes); if (rc) @@ -4339,8 +4338,7 @@ smb2_async_writev(struct cifs_writedata *wdata, if (wdata->credits.value > 0) { shdr->CreditCharge = cpu_to_le16(DIV_ROUND_UP(wdata->bytes, SMB2_MAX_BUFFER_SIZE)); - shdr->CreditRequest = - cpu_to_le16(le16_to_cpu(shdr->CreditCharge) + 1); + shdr->CreditRequest = cpu_to_le16(le16_to_cpu(shdr->CreditCharge) + 8);
rc = adjust_credits(server, &wdata->credits, wdata->bytes); if (rc)