[PATCH v2] net: ll_temac: Use one return statement instead of two

[PATCH v2] net: ll_temac: Use one return statement instead of two

[Michal Simek]

Use one return statement instead of two to simplify the code.
Both are returning the same value.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
---

Changes in v2:
- Use the same SoB as sender address - Reported-by: Julia Lawall

 drivers/net/ethernet/xilinx/ll_temac_main.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/xilinx/ll_temac_main.c b/drivers/net/ethernet/xilinx/ll_temac_main.c
index 062483f334a5..cfb6bdb37fdc 100644
--- a/drivers/net/ethernet/xilinx/ll_temac_main.c
+++ b/drivers/net/ethernet/xilinx/ll_temac_main.c
@@ -688,10 +688,8 @@ static int temac_start_xmit(struct sk_buff *skb, struct net_device *ndev)
 	cur_p = &lp->tx_bd_v[lp->tx_bd_tail];
 
 	if (temac_check_tx_bd_space(lp, num_frag)) {
-		if (!netif_queue_stopped(ndev)) {
+		if (!netif_queue_stopped(ndev))
 			netif_stop_queue(ndev);
-			return NETDEV_TX_BUSY;
-		}
 		return NETDEV_TX_BUSY;
 	}
 
-- 
2.3.5

[PATCH v2] net: ll_temac: Use one return statement instead of two

[Michal Simek]

Use one return statement instead of two to simplify the code.
Both are returning the same value.

Signed-off-by: Michal Simek <michal.simek@xilinx.com>
---

Changes in v2:
- Use the same SoB as sender address - Reported-by: Julia Lawall

 drivers/net/ethernet/xilinx/ll_temac_main.c | 4 +---
 1 file changed, 1 insertion(+), 3 deletions(-)

diff --git a/drivers/net/ethernet/xilinx/ll_temac_main.c b/drivers/net/ethernet/xilinx/ll_temac_main.c
index 062483f334a5..cfb6bdb37fdc 100644
--- a/drivers/net/ethernet/xilinx/ll_temac_main.c
+++ b/drivers/net/ethernet/xilinx/ll_temac_main.c
@@ -688,10 +688,8 @@ static int temac_start_xmit(struct sk_buff *skb, struct net_device *ndev)
 	cur_p = &lp->tx_bd_v[lp->tx_bd_tail];
 
 	if (temac_check_tx_bd_space(lp, num_frag)) {
-		if (!netif_queue_stopped(ndev)) {
+		if (!netif_queue_stopped(ndev))
 			netif_stop_queue(ndev);
-			return NETDEV_TX_BUSY;
-		}
 		return NETDEV_TX_BUSY;
 	}
 
-- 
2.3.5

Re: [PATCH v2] net: ll_temac: Use one return statement instead of two

[Joe Perches]

On Mon, 2015-05-11 at 16:05 +0200, Michal Simek wrote:
> Use one return statement instead of two to simplify the code.
> Both are returning the same value.

trivia and FYI:

I think there are about 100 of these in kernel tree
and I'm not going to submit patches.

$ grep-2.5.4 -rP --include=*.[ch] -n "^([\t]+)\treturn[ \t]+([A-Za-z0-9\_\>\(\)\.\>\[\]\-]+);\n(\1}\n)?\1return[ \t]+\2;" * | \
  grep -P "^[\w\/\.]+:\d+:"
arch/x86/kvm/emulate.c:2122:		return rc;
arch/um/kernel/exitcode.c:75:		return 0;
arch/um/drivers/mconsole_kern.c:784:		return 0;
arch/powerpc/include/asm/floppy.h:84:		return IRQ_HANDLED;
drivers/gpu/drm/via/via_video.c:90:		return 0;
drivers/gpu/drm/via/via_verifier.c:514:		return 2;
drivers/gpu/drm/omapdrm/omap_fb.c:375:		return fb;
drivers/gpu/drm/radeon/evergreen_cs.c:312:		return -EINVAL;
drivers/gpu/drm/radeon/ni.c:2161:		return r;
drivers/isdn/capi/capidrv.c:1682:		return -EINVAL;
drivers/isdn/hisax/saphir.c:237:		return (0);
drivers/isdn/hisax/teles3.c:249:		return (0);
drivers/isdn/hisax/hfc_pci.c:1628:		return (0);
drivers/isdn/hisax/ix1_micro.c:206:		return (0);
drivers/isdn/hisax/avm_a1p.c:211:		return 0;
drivers/isdn/hisax/w6692.c:987:		return (0);
drivers/isdn/hisax/asuscom.c:292:		return (0);
drivers/isdn/hisax/avm_a1.c:175:		return (0);
drivers/isdn/hisax/hfcscard.c:133:		return (0);
drivers/isdn/hisax/teleint.c:257:		return (0);
drivers/isdn/hisax/hfc_sx.c:1379:		return (0);
drivers/isdn/hisax/teles0.c:261:		return (0);
drivers/isdn/hisax/mic.c:185:		return (0);
drivers/isdn/hisax/nj_u.c:126:		return (0);
drivers/isdn/hisax/bkm_a4t.c:251:		return (0);
drivers/isdn/hisax/avm_pci.c:716:		return (0);
drivers/isdn/hisax/bkm_a8.c:253:		return (0);
drivers/isdn/hisax/sportster.c:181:		return (0);
drivers/isdn/hisax/niccy.c:220:		return 0;
drivers/isdn/hisax/telespci.c:281:		return (0);
drivers/isdn/hisax/isurf.c:170:		return (0);
drivers/isdn/hisax/gazel.c:427:		return (0);
drivers/isdn/hisax/nj_s.c:146:		return (0);
drivers/isdn/i4l/isdn_v110.c:613:		return 0;
drivers/isdn/hardware/avm/b1.c:634:		return IRQ_HANDLED;
drivers/staging/lustre/lustre/osc/osc_request.c:2429:		return rc;
drivers/staging/dgap/dgap.c:421:		return 0;
drivers/staging/comedi/drivers/ni_mio_common.c:4997:			return 0;
drivers/staging/comedi/drivers/usbduxsigma.c:1245:		return -EINVAL;
drivers/staging/comedi/drivers/cb_pcidas64.c:1961:		return -EINVAL;
drivers/staging/comedi/drivers/usbdux.c:1384:		return -EINVAL;
drivers/hid/wacom_wac.c:1063:		return 0;
drivers/crypto/qat/qat_dh895xcc/adf_isr.c:144:		return ret;
drivers/crypto/qat/qat_dh895xcc/adf_dh895xcc_hw_data.c:135:		return DEV_SKU_UNKNOWN;
drivers/parisc/led.c:168:		return 0;
drivers/block/drbd/drbd_receiver.c:3457:		return tfm;
drivers/iio/light/tcs3472.c:180:		return -EINVAL;
drivers/misc/genwqe/card_base.c:1335:		return 0;
drivers/mfd/tps80031.c:267:		return ret;
drivers/mfd/tps80031.c:307:		return ret;
drivers/mfd/palmas.c:374:		return ret;
drivers/char/tpm/tpm_infineon.c:308:		return -EIO;
drivers/scsi/wd7000.c:1558:		return FAILED;
drivers/scsi/scsi_error.c:720:		return FAILED;
drivers/scsi/scsi_error.c:1920:		return FAILED;
drivers/scsi/be2iscsi/be_mgmt.c:1414:		return rc;
drivers/scsi/initio.c:2480:		return host->phase;
drivers/scsi/qla2xxx/qla_mr.c:692:		return str;
drivers/scsi/bfa/bfa_fcbuild.c:208:		return FC_PARSE_OK;
drivers/scsi/arcmsr/arcmsr_hba.c:3736:		return rtnval;
drivers/scsi/BusLogic.c:3317:		return SUCCESS;
drivers/net/ethernet/xilinx/ll_temac_main.c:693:			return NETDEV_TX_BUSY;
drivers/net/ethernet/qlogic/qlge/qlge_main.c:3251:		return err;
drivers/net/ethernet/qlogic/qlge/qlge_main.c:3295:		return err;
drivers/net/ethernet/qlogic/qlge/qlge_main.c:3638:		return status;
drivers/net/ethernet/intel/ixgbe/ixgbe_phy.c:1492:		return 0;
drivers/net/wan/x25_asy.c:350:		return NETDEV_TX_OK;
drivers/net/wireless/rtlwifi/rtl8192cu/mac.c:227:		return rst;
drivers/net/wireless/brcm80211/brcmfmac/cfg80211.c:1176:		return err;
drivers/net/plip/plip.c:739:		return OK;
drivers/net/plip/plip.c:853:		return OK;
drivers/net/caif/caif_serial.c:411:		return result;
drivers/spi/spi.c:1057:		return ret;
drivers/usb/gadget/function/f_loopback.c:560:		return ret;
drivers/infiniband/ulp/isert/ib_isert.c:2143:		return ret;
drivers/infiniband/hw/ocrdma/ocrdma_verbs.c:1808:		return status;
drivers/infiniband/hw/ocrdma/ocrdma_hw.c:151:		return IB_QPS_ERR;
drivers/infiniband/hw/ocrdma/ocrdma_hw.c:170:		return OCRDMA_QPS_ERR;
drivers/infiniband/hw/cxgb4/iw_cxgb4.h:618:		return IB_QPS_ERR;
drivers/infiniband/hw/mlx5/qp.c:1547:		return 0;
drivers/media/pci/smipcie/smipcie.c:261:		return ret;
drivers/media/i2c/tw2804.c:196:		return 0;
drivers/video/fbdev/68328fb.c:355:		return 0;
drivers/video/fbdev/vfb.c:378:		return 0;
fs/namespace.c:1935:			return mp;
kernel/auditsc.c:425:		return 0;
kernel/trace/trace_uprobe.c:1192:		return 0;
net/sctp/outqueue.c:150:		return 0;
net/caif/cfctrl.c:112:		return false;
net/ipv4/netfilter/nf_nat_snmp_basic.c:1299:		return ret;
net/tipc/server.c:612:		return ret;
sound/pci/riptide/riptide.c:2003:		return err;
sound/pci/rme9652/hdsp.c:2711:		return 0;
sound/core/compress_offload.c:879:		return ret;
sound/core/seq/oss/seq_oss_timer.c:274:		return 0;
sound/oss/sb_midi.c:100:		return 1;
sound/soc/codecs/max98925.c:120:		return 0;
tools/perf/util/config.c:391:		return 0;
tools/perf/util/alias.c:13:		return 0;

Re: [PATCH v2] net: ll_temac: Use one return statement instead of two

[Michal Simek]

On 05/11/2015 04:42 PM, Joe Perches wrote:
> On Mon, 2015-05-11 at 16:05 +0200, Michal Simek wrote:
>> Use one return statement instead of two to simplify the code.
>> Both are returning the same value.
> 
> trivia and FYI:

But still correct right?

> 
> I think there are about 100 of these in kernel tree
> and I'm not going to submit patches.
> 
> $ grep-2.5.4 -rP --include=*.[ch] -n "^([\t]+)\treturn[ \t]+([A-Za-z0-9\_\>\(\)\.\>\[\]\-]+);\n(\1}\n)?\1return[ \t]+\2;" * | \
>   grep -P "^[\w\/\.]+:\d+:"

Are you suggesting that someone else should send patches for it?
I expect that this is something what Julia can check with coccinelle
and can be added to scripts folder.

Thanks,
Michal

Re: [PATCH v2] net: ll_temac: Use one return statement instead of two

[Joe Perches]

On Mon, 2015-05-11 at 17:26 +0200, Michal Simek wrote:
> On 05/11/2015 04:42 PM, Joe Perches wrote:
> > On Mon, 2015-05-11 at 16:05 +0200, Michal Simek wrote:
> >> Use one return statement instead of two to simplify the code.
> >> Both are returning the same value.
> > 
> > trivia and FYI:
> 
> But still correct right?

Yes.

> > I think there are about 100 of these in kernel tree
> > and I'm not going to submit patches.
> > 
> > $ grep-2.5.4 -rP --include=*.[ch] -n "^([\t]+)\treturn[ \t]+([A-Za-z0-9\_\>\(\)\.\>\[\]\-]+);\n(\1}\n)?\1return[ \t]+\2;" * | \
> >   grep -P "^[\w\/\.]+:\d+:"
> 
> Are you suggesting that someone else should send patches for it?

I get enough grief for doing style oriented patches,
so if someone else wants to, sure.

> I expect that this is something what Julia can check with coccinelle
> and can be added to scripts folder.

Maybe true.

A coccinelle script might be rather more complicated
than the simpler grep above, but perhaps the script
could be a bit more complete as it could likely look
at more code indentation styles.

Re: [PATCH v2] net: ll_temac: Use one return statement instead of two

[Julia Lawall]

On Mon, 11 May 2015, Michal Simek wrote:

> On 05/11/2015 04:42 PM, Joe Perches wrote:
> > On Mon, 2015-05-11 at 16:05 +0200, Michal Simek wrote:
> >> Use one return statement instead of two to simplify the code.
> >> Both are returning the same value.
> >
> > trivia and FYI:
>
> But still correct right?
>
> >
> > I think there are about 100 of these in kernel tree
> > and I'm not going to submit patches.
> >
> > $ grep-2.5.4 -rP --include=*.[ch] -n "^([\t]+)\treturn[ \t]+([A-Za-z0-9\_\>\(\)\.\>\[\]\-]+);\n(\1}\n)?\1return[ \t]+\2;" * | \
> >   grep -P "^[\w\/\.]+:\d+:"
>
> Are you suggesting that someone else should send patches for it?
> I expect that this is something what Julia can check with coccinelle
> and can be added to scripts folder.

I can do something based on the original patch.  I have no idea what the
regular expression does :)

julia

Re: [PATCH v2] net: ll_temac: Use one return statement instead of two

[Michal Simek]

On 05/11/2015 05:39 PM, Joe Perches wrote:
> On Mon, 2015-05-11 at 17:26 +0200, Michal Simek wrote:
>> On 05/11/2015 04:42 PM, Joe Perches wrote:
>>> On Mon, 2015-05-11 at 16:05 +0200, Michal Simek wrote:
>>>> Use one return statement instead of two to simplify the code.
>>>> Both are returning the same value.
>>>
>>> trivia and FYI:
>>
>> But still correct right?
> 
> Yes.

Does this mean that you have added me your ACK?

>>> I think there are about 100 of these in kernel tree
>>> and I'm not going to submit patches.
>>>
>>> $ grep-2.5.4 -rP --include=*.[ch] -n "^([\t]+)\treturn[ \t]+([A-Za-z0-9\_\>\(\)\.\>\[\]\-]+);\n(\1}\n)?\1return[ \t]+\2;" * | \
>>>   grep -P "^[\w\/\.]+:\d+:"
>>
>> Are you suggesting that someone else should send patches for it?
> 
> I get enough grief for doing style oriented patches,
> so if someone else wants to, sure.

yep - the only reason I have submitted this was that this is the diff in
the xilinx tree compare to mainline and I don't want to simple revert it
because it is correct.

>> I expect that this is something what Julia can check with coccinelle
>> and can be added to scripts folder.
> 
> Maybe true.
> 
> A coccinelle script might be rather more complicated
> than the simpler grep above, but perhaps the script
> could be a bit more complete as it could likely look
> at more code indentation styles.

Julia: Any comment?

Thanks,
Michal

Re: [PATCH v2] net: ll_temac: Use one return statement instead of two

[Julia Lawall]

> > A coccinelle script might be rather more complicated
> > than the simpler grep above, but perhaps the script
> > could be a bit more complete as it could likely look
> > at more code indentation styles.
>
> Julia: Any comment?

Here is what I had in mind:

if (...) {
  ... when != goto l;
  return C;
}
return C;

C is a constant, to avoid that its value depends on the code in the ...

julia

Re: [PATCH v2] net: ll_temac: Use one return statement instead of two

[Joe Perches]

On Mon, 2015-05-11 at 17:48 +0200, Julia Lawall wrote:
> > > A coccinelle script might be rather more complicated
> > > than the simpler grep above, but perhaps the script
> > > could be a bit more complete as it could likely look
> > > at more code indentation styles.
> >
> > Julia: Any comment?
> 
> Here is what I had in mind:
> 
> if (...) {
>   ... when != goto l;
>   return C;
> }
> return C;
> 
> C is a constant, to avoid that its value depends on the code in the ...

Sure but I think that would miss several instances like:

	switch (<foo>) {
		...
	default:
		return <bar>;
	}
	return <bar>;

or the similar

	if (foo) {
		if (qux)
			return <bar>;
	} else {
		return <baz>;
	}

	return <baz>;

Re: [PATCH v2] net: ll_temac: Use one return statement instead of two

[Joe Perches]

On Mon, 2015-05-11 at 17:44 +0200, Michal Simek wrote:
> On 05/11/2015 05:39 PM, Joe Perches wrote:
> > On Mon, 2015-05-11 at 17:26 +0200, Michal Simek wrote:
> >> On 05/11/2015 04:42 PM, Joe Perches wrote:
> >>> On Mon, 2015-05-11 at 16:05 +0200, Michal Simek wrote:
> >>>> Use one return statement instead of two to simplify the code.
> >>>> Both are returning the same value.
> >>>
> >>> trivia and FYI:
> >>
> >> But still correct right?
> > 
> > Yes.
> 
> Does this mean that you have added me your ACK?

No,

I don't generally ack patches as I think it's
not very useful.

Re: [PATCH v2] net: ll_temac: Use one return statement instead of two

[Michal Simek]

On 05/11/2015 06:10 PM, Joe Perches wrote:
> On Mon, 2015-05-11 at 17:44 +0200, Michal Simek wrote:
>> On 05/11/2015 05:39 PM, Joe Perches wrote:
>>> On Mon, 2015-05-11 at 17:26 +0200, Michal Simek wrote:
>>>> On 05/11/2015 04:42 PM, Joe Perches wrote:
>>>>> On Mon, 2015-05-11 at 16:05 +0200, Michal Simek wrote:
>>>>>> Use one return statement instead of two to simplify the code.
>>>>>> Both are returning the same value.
>>>>>
>>>>> trivia and FYI:
>>>>
>>>> But still correct right?
>>>
>>> Yes.
>>
>> Does this mean that you have added me your ACK?
> 
> No,
> 
> I don't generally ack patches as I think it's
> not very useful.

OK. Fair enough.

Thanks,
Michal

Re: [PATCH v2] net: ll_temac: Use one return statement instead of two

[David Miller]

From: Michal Simek <michal.simek@xilinx.com>
Date: Mon, 11 May 2015 16:05:02 +0200

> Use one return statement instead of two to simplify the code.
> Both are returning the same value.
> 
> Signed-off-by: Michal Simek <michal.simek@xilinx.com>

Applied, thanks.

[PATCH v2] net: ll_temac: Use one return statement instead of two

[David Miller]

From: Michal Simek <michal.simek@xilinx.com>
Date: Mon, 11 May 2015 16:05:02 +0200

> Use one return statement instead of two to simplify the code.
> Both are returning the same value.
> 
> Signed-off-by: Michal Simek <michal.simek@xilinx.com>

Applied, thanks.

Re: [PATCH v2] net: ll_temac: Use one return statement instead of two

[Julia Lawall]

On Mon, 11 May 2015, Joe Perches wrote:

> On Mon, 2015-05-11 at 17:48 +0200, Julia Lawall wrote:
> > > > A coccinelle script might be rather more complicated
> > > > than the simpler grep above, but perhaps the script
> > > > could be a bit more complete as it could likely look
> > > > at more code indentation styles.
> > >
> > > Julia: Any comment?
> > 
> > Here is what I had in mind:
> > 
> > if (...) {
> >   ... when != goto l;
> >   return C;
> > }
> > return C;
> > 
> > C is a constant, to avoid that its value depends on the code in the ...
> 
> Sure but I think that would miss several instances like:
> 
> 	switch (<foo>) {
> 		...
> 	default:
> 		return <bar>;
> 	}
> 	return <bar>;

Switch Coccinelle is not very good at...

> or the similar
> 
> 	if (foo) {
> 		if (qux)
> 			return <bar>;
> 	} else {
> 		return <baz>;
> 	}
> 
> 	return <baz>;

It seems improbable, but I could look for that.  Unfortunately, I don't 
see a way to deal with arbitrarily nested ifs.  Basically, the control 
flow from one return doesn't go to the other.  It goes from the return to 
the outside of the function.  I guess something could be done by renaming 
all of the returns to function calls, but that tends to make a mess.  It 
could be done to see if such cases are worth considering though.

Another similar and popular construction is:

if (...) {
  ...
  goto l;
}
l:

julia