| Message ID | 20210319040447.527-1-liew.s.piaw@gmail.com (mailing list archive) |
|---|---|
| State | Changes Requested |
| Delegated to: | Netdev Maintainers |
| Headers | show |
| Series | atl1c: optimize rx loop | expand |
| Context | Check | Description |
|---|---|---|
| netdev/cover_letter | success | Link |
| netdev/fixes_present | success | Link |
| netdev/patch_count | success | Link |
| netdev/tree_selection | success | Guessed tree name to be net-next |
| netdev/subject_prefix | warning | Target tree name not specified in the subject |
| netdev/cc_maintainers | warning | 5 maintainers not CCed: zhangchangzhong@huawei.com christophe.jaillet@wanadoo.fr tglx@linutronix.de yanaijie@huawei.com jesse.brandeburg@intel.com |
| netdev/source_inline | success | Was 0 now: 0 |
| netdev/verify_signedoff | success | Link |
| netdev/module_param | success | Was 0 now: 0 |
| netdev/build_32bit | success | Errors and warnings before: 31 this patch: 31 |
| netdev/kdoc | success | Errors and warnings before: 3 this patch: 3 |
| netdev/verify_fixes | success | Link |
| netdev/checkpatch | success | total: 0 errors, 0 warnings, 0 checks, 10 lines checked |
| netdev/build_allmodconfig_warn | success | Errors and warnings before: 31 this patch: 31 |
| netdev/header_inline | success | Link |
On Fri, Mar 19, 2021 at 12:04:47PM +0800, Sieng Piaw Liew wrote: > Remove this trivial bit of inefficiency from the rx receive loop, > results in increase of a few Mbps in iperf3. Tested on Intel Core2 > platform. > > Signed-off-by: Sieng Piaw Liew <liew.s.piaw@gmail.com> > --- > drivers/net/ethernet/atheros/atl1c/atl1c_main.c | 4 +--- > 1 file changed, 1 insertion(+), 3 deletions(-) > > diff --git a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c > index 3f65f2b370c5..b995f9a0479c 100644 > --- a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c > +++ b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c > @@ -1796,9 +1796,7 @@ static void atl1c_clean_rx_irq(struct atl1c_adapter *adapter, > struct atl1c_recv_ret_status *rrs; > struct atl1c_buffer *buffer_info; > > - while (1) { > - if (*work_done >= work_to_do) > - break; > + while (*work_done < work_to_do) { It should not change anything, or only based on the compiler's optimization and should not result in a measurable difference because what it does is exactly the same. Have you really compared the compiled output code to explain the difference ? I strongly suspect you'll find no difference at all. Thus for me it's certainly not an optimization, it could be qualified as a cleanup to improve code readability however. Willy
On Fri, Mar 19, 2021 at 05:15:35AM +0100, Willy Tarreau wrote: > On Fri, Mar 19, 2021 at 12:04:47PM +0800, Sieng Piaw Liew wrote: > > Remove this trivial bit of inefficiency from the rx receive loop, > > results in increase of a few Mbps in iperf3. Tested on Intel Core2 > > platform. > > > > Signed-off-by: Sieng Piaw Liew <liew.s.piaw@gmail.com> > > --- > > drivers/net/ethernet/atheros/atl1c/atl1c_main.c | 4 +--- > > 1 file changed, 1 insertion(+), 3 deletions(-) > > > > diff --git a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c > > index 3f65f2b370c5..b995f9a0479c 100644 > > --- a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c > > +++ b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c > > @@ -1796,9 +1796,7 @@ static void atl1c_clean_rx_irq(struct atl1c_adapter *adapter, > > struct atl1c_recv_ret_status *rrs; > > struct atl1c_buffer *buffer_info; > > > > - while (1) { > > - if (*work_done >= work_to_do) > > - break; > > + while (*work_done < work_to_do) { > > It should not change anything, or only based on the compiler's optimization > and should not result in a measurable difference because what it does is > exactly the same. Have you really compared the compiled output code to > explain the difference ? I strongly suspect you'll find no difference at > all. > > Thus for me it's certainly not an optimization, it could be qualified as > a cleanup to improve code readability however. > > Willy You're right. Objdump and diff showed no difference. Regards, Sieng Piaw
diff --git a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c index 3f65f2b370c5..b995f9a0479c 100644 --- a/drivers/net/ethernet/atheros/atl1c/atl1c_main.c +++ b/drivers/net/ethernet/atheros/atl1c/atl1c_main.c @@ -1796,9 +1796,7 @@ static void atl1c_clean_rx_irq(struct atl1c_adapter *adapter, struct atl1c_recv_ret_status *rrs; struct atl1c_buffer *buffer_info; - while (1) { - if (*work_done >= work_to_do) - break; + while (*work_done < work_to_do) { rrs = ATL1C_RRD_DESC(rrd_ring, rrd_ring->next_to_clean); if (likely(RRS_RXD_IS_VALID(rrs->word3))) { rfd_num = (rrs->word0 >> RRS_RX_RFD_CNT_SHIFT) &
Remove this trivial bit of inefficiency from the rx receive loop, results in increase of a few Mbps in iperf3. Tested on Intel Core2 platform. Signed-off-by: Sieng Piaw Liew <liew.s.piaw@gmail.com> --- drivers/net/ethernet/atheros/atl1c/atl1c_main.c | 4 +--- 1 file changed, 1 insertion(+), 3 deletions(-)