RE: [PATCH 3/4] scsi: storvsc: Refactor the code in storvsc_channel_init()

[KY Srinivasan]

> -Original Message-
> From: Dan Carpenter [mailto:dan.carpen...@oracle.com]
> Sent: Friday, December 11, 2015 2:41 AM
> To: KY Srinivasan 
> Cc: gre...@linuxfoundation.org; linux-kernel@vger.kernel.org;
> de...@linuxdriverproject.org; oher...@suse.com;
> jbottom...@parallels.com; h...@infradead.org; linux-s...@vger.kernel.org;
> a...@canonical.com; vkuzn...@redhat.com; jasow...@redhat.com;
> martin.peter...@oracle.com
> Subject: Re: [PATCH 3/4] scsi: storvsc: Refactor the code in
> storvsc_channel_init()
> 
> On Thu, Dec 10, 2015 at 04:14:19PM -0800, K. Y. Srinivasan wrote:
> > @@ -753,27 +740,62 @@ static int storvsc_channel_init(struct hv_device
> *device, bool is_fc)
> >VM_PKT_DATA_INBAND,
> >
> VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
> > if (ret != 0)
> > -   goto cleanup;
> > +   goto done;
> >
> > t = wait_for_completion_timeout(>wait_event, 5*HZ);
> > if (t == 0) {
> > ret = -ETIMEDOUT;
> > -   goto cleanup;
> > +   goto done;
> > }
> >
> > +   if (!status_check)
> > +   goto done;
> 
> See?  This goto looks exactly the same as the earlier buggy goto but
> it's actually correct.  Meanwhile if you just used an explicit
> "return 0;" then it would be easy to understand.
> 
> I rant about this all the time but it's because it's bad deliberately.
> It's normal to have bugs, but this deliberate stuff really I can't
> understand it...
> 
> > +
> > if (vstor_packet->operation != VSTOR_OPERATION_COMPLETE_IO
> ||
> > vstor_packet->status != 0) {
> > ret = -EINVAL;
> > -   goto cleanup;
> > +   goto done;
> > }
> >
> > +done:
> > +   return ret;
> > +}
> > +
> > +static int storvsc_channel_init(struct hv_device *device, bool is_fc)
> > +{
> > +   struct storvsc_device *stor_device;
> > +   struct storvsc_cmd_request *request;
> > +   struct vstor_packet *vstor_packet;
> > +   int ret, i;
> > +   int max_chns;
> > +   bool process_sub_channels = false;
> > +
> > +   stor_device = get_out_stor_device(device);
> > +   if (!stor_device)
> > +   return -ENODEV;
> > +
> > +   request = _device->init_request;
> > +   vstor_packet = >vstor_packet;
> > +
> > +   /*
> > +* Now, initiate the vsc/vsp initialization protocol on the open
> > +* channel
> > +*/
> > +   memset(request, 0, sizeof(struct storvsc_cmd_request));
> > +   vstor_packet->operation =
> VSTOR_OPERATION_BEGIN_INITIALIZATION;
> > +   ret = storvsc_execute_vstor_op(device, request, true);
> > +   if (ret)
> > +   goto cleanup;
> 
> 10 lines earlier there is an explicit "return -ENODEV" so it's not as if
> writing explicit returns will kill you.

Thanks Dan; I will cleanup the code and resend.

Regards,

K. Y



--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Re: [PATCH 3/4] scsi: storvsc: Refactor the code in storvsc_channel_init()

[Dan Carpenter]

On Thu, Dec 10, 2015 at 04:14:19PM -0800, K. Y. Srinivasan wrote:
> @@ -753,27 +740,62 @@ static int storvsc_channel_init(struct hv_device 
> *device, bool is_fc)
>  VM_PKT_DATA_INBAND,
>  VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
>   if (ret != 0)
> - goto cleanup;
> + goto done;
>  
>   t = wait_for_completion_timeout(>wait_event, 5*HZ);
>   if (t == 0) {
>   ret = -ETIMEDOUT;
> - goto cleanup;
> + goto done;
>   }
>  
> + if (!status_check)
> + goto done;

See?  This goto looks exactly the same as the earlier buggy goto but
it's actually correct.  Meanwhile if you just used an explicit
"return 0;" then it would be easy to understand.

I rant about this all the time but it's because it's bad deliberately.
It's normal to have bugs, but this deliberate stuff really I can't
understand it...

> +
>   if (vstor_packet->operation != VSTOR_OPERATION_COMPLETE_IO ||
>   vstor_packet->status != 0) {
>   ret = -EINVAL;
> - goto cleanup;
> + goto done;
>   }
>  
> +done:
> + return ret;
> +}
> +
> +static int storvsc_channel_init(struct hv_device *device, bool is_fc)
> +{
> + struct storvsc_device *stor_device;
> + struct storvsc_cmd_request *request;
> + struct vstor_packet *vstor_packet;
> + int ret, i;
> + int max_chns;
> + bool process_sub_channels = false;
> +
> + stor_device = get_out_stor_device(device);
> + if (!stor_device)
> + return -ENODEV;
> +
> + request = _device->init_request;
> + vstor_packet = >vstor_packet;
> +
> + /*
> +  * Now, initiate the vsc/vsp initialization protocol on the open
> +  * channel
> +  */
> + memset(request, 0, sizeof(struct storvsc_cmd_request));
> + vstor_packet->operation = VSTOR_OPERATION_BEGIN_INITIALIZATION;
> + ret = storvsc_execute_vstor_op(device, request, true);
> + if (ret)
> + goto cleanup;

10 lines earlier there is an explicit "return -ENODEV" so it's not as if
writing explicit returns will kill you.

regards,
dan carpenter

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Re: [PATCH 3/4] scsi: storvsc: Refactor the code in storvsc_channel_init()

[Johannes Thumshirn]

On Thu, Dec 10, 2015 at 04:14:19PM -0800, K. Y. Srinivasan wrote:
> The function storvsc_channel_init() repeatedly interacts with the host to
> extract various channel properties. Refactor this code to eliminate code
> repetition.
> 
> Signed-off-by: K. Y. Srinivasan 
> Reviewed-by: Long Li 
> Tested-by: Alex Ng 
> ---
>  drivers/scsi/storvsc_drv.c |  155 
> 
>  1 files changed, 57 insertions(+), 98 deletions(-)
> 
> diff --git a/drivers/scsi/storvsc_drv.c b/drivers/scsi/storvsc_drv.c
> index b94d471..6f18e94 100644
> --- a/drivers/scsi/storvsc_drv.c
> +++ b/drivers/scsi/storvsc_drv.c
> @@ -721,29 +721,16 @@ static void cache_wwn(struct storvsc_device 
> *stor_device,
>   }
>  }
>  
> -static int storvsc_channel_init(struct hv_device *device, bool is_fc)
> +static int storvsc_execute_vstor_op(struct hv_device *device,
> + struct storvsc_cmd_request *request,
> + bool status_check)
>  {
> - struct storvsc_device *stor_device;
> - struct storvsc_cmd_request *request;
>   struct vstor_packet *vstor_packet;
> - int ret, t, i;
> - int max_chns;
> - bool process_sub_channels = false;
> -
> - stor_device = get_out_stor_device(device);
> - if (!stor_device)
> - return -ENODEV;
> + int ret, t;
>  
> - request = _device->init_request;
>   vstor_packet = >vstor_packet;
>  
> - /*
> -  * Now, initiate the vsc/vsp initialization protocol on the open
> -  * channel
> -  */
> - memset(request, 0, sizeof(struct storvsc_cmd_request));
>   init_completion(>wait_event);
> - vstor_packet->operation = VSTOR_OPERATION_BEGIN_INITIALIZATION;
>   vstor_packet->flags = REQUEST_COMPLETION_FLAG;
>  
>   ret = vmbus_sendpacket(device->channel, vstor_packet,
> @@ -753,27 +740,62 @@ static int storvsc_channel_init(struct hv_device 
> *device, bool is_fc)
>  VM_PKT_DATA_INBAND,
>  VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
>   if (ret != 0)
> - goto cleanup;
> + goto done;
>  
>   t = wait_for_completion_timeout(>wait_event, 5*HZ);
>   if (t == 0) {
>   ret = -ETIMEDOUT;
> - goto cleanup;
> + goto done;
>   }
>  
> + if (!status_check)
> + goto done;
> +
>   if (vstor_packet->operation != VSTOR_OPERATION_COMPLETE_IO ||
>   vstor_packet->status != 0) {
>   ret = -EINVAL;
> - goto cleanup;
> + goto done;
>   }
>  
> +done:
> + return ret;
> +}
> +
> +static int storvsc_channel_init(struct hv_device *device, bool is_fc)
> +{
> + struct storvsc_device *stor_device;
> + struct storvsc_cmd_request *request;
> + struct vstor_packet *vstor_packet;
> + int ret, i;
> + int max_chns;
> + bool process_sub_channels = false;
> +
> + stor_device = get_out_stor_device(device);
> + if (!stor_device)
> + return -ENODEV;
> +
> + request = _device->init_request;
> + vstor_packet = >vstor_packet;
> +
> + /*
> +  * Now, initiate the vsc/vsp initialization protocol on the open
> +  * channel
> +  */
> + memset(request, 0, sizeof(struct storvsc_cmd_request));
> + vstor_packet->operation = VSTOR_OPERATION_BEGIN_INITIALIZATION;
> + ret = storvsc_execute_vstor_op(device, request, true);
> + if (ret)
> + goto cleanup;
> +
> + /*
> +  * Query host supported protocol version.
> +  */
>  
>   for (i = 0; i < ARRAY_SIZE(vmstor_protocols); i++) {
>   /* reuse the packet for version range supported */
>   memset(vstor_packet, 0, sizeof(struct vstor_packet));
>   vstor_packet->operation =
>   VSTOR_OPERATION_QUERY_PROTOCOL_VERSION;
> - vstor_packet->flags = REQUEST_COMPLETION_FLAG;
>  
>   vstor_packet->version.major_minor =
>   vmstor_protocols[i].protocol_version;
> @@ -783,20 +805,9 @@ static int storvsc_channel_init(struct hv_device 
> *device, bool is_fc)
>*/
>   vstor_packet->version.revision = 0;
>  
> - ret = vmbus_sendpacket(device->channel, vstor_packet,
> -(sizeof(struct vstor_packet) -
> - vmscsi_size_delta),
> -(unsigned long)request,
> -VM_PKT_DATA_INBAND,
> -VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
> - if (ret != 0)
> - goto cleanup;
> -
> - t = wait_for_completion_timeout(>wait_event, 5*HZ);
> - if (t == 0) {
> - ret = -ETIMEDOUT;
> + ret = storvsc_execute_vstor_op(device, request, false);
> + if (ret)
>   goto cleanup;
> - }
>  
>   

Re: [PATCH 3/4] scsi: storvsc: Refactor the code in storvsc_channel_init()

[Dan Carpenter]

On Thu, Dec 10, 2015 at 04:14:19PM -0800, K. Y. Srinivasan wrote:
> @@ -753,27 +740,62 @@ static int storvsc_channel_init(struct hv_device 
> *device, bool is_fc)
>  VM_PKT_DATA_INBAND,
>  VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
>   if (ret != 0)
> - goto cleanup;
> + goto done;
>  
>   t = wait_for_completion_timeout(>wait_event, 5*HZ);
>   if (t == 0) {
>   ret = -ETIMEDOUT;
> - goto cleanup;
> + goto done;
>   }
>  
> + if (!status_check)
> + goto done;

See?  This goto looks exactly the same as the earlier buggy goto but
it's actually correct.  Meanwhile if you just used an explicit
"return 0;" then it would be easy to understand.

I rant about this all the time but it's because it's bad deliberately.
It's normal to have bugs, but this deliberate stuff really I can't
understand it...

> +
>   if (vstor_packet->operation != VSTOR_OPERATION_COMPLETE_IO ||
>   vstor_packet->status != 0) {
>   ret = -EINVAL;
> - goto cleanup;
> + goto done;
>   }
>  
> +done:
> + return ret;
> +}
> +
> +static int storvsc_channel_init(struct hv_device *device, bool is_fc)
> +{
> + struct storvsc_device *stor_device;
> + struct storvsc_cmd_request *request;
> + struct vstor_packet *vstor_packet;
> + int ret, i;
> + int max_chns;
> + bool process_sub_channels = false;
> +
> + stor_device = get_out_stor_device(device);
> + if (!stor_device)
> + return -ENODEV;
> +
> + request = _device->init_request;
> + vstor_packet = >vstor_packet;
> +
> + /*
> +  * Now, initiate the vsc/vsp initialization protocol on the open
> +  * channel
> +  */
> + memset(request, 0, sizeof(struct storvsc_cmd_request));
> + vstor_packet->operation = VSTOR_OPERATION_BEGIN_INITIALIZATION;
> + ret = storvsc_execute_vstor_op(device, request, true);
> + if (ret)
> + goto cleanup;

10 lines earlier there is an explicit "return -ENODEV" so it's not as if
writing explicit returns will kill you.

regards,
dan carpenter

--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/

Re: [PATCH 3/4] scsi: storvsc: Refactor the code in storvsc_channel_init()

[Johannes Thumshirn]

On Thu, Dec 10, 2015 at 04:14:19PM -0800, K. Y. Srinivasan wrote:
> The function storvsc_channel_init() repeatedly interacts with the host to
> extract various channel properties. Refactor this code to eliminate code
> repetition.
> 
> Signed-off-by: K. Y. Srinivasan 
> Reviewed-by: Long Li 
> Tested-by: Alex Ng 
> ---
>  drivers/scsi/storvsc_drv.c |  155 
> 
>  1 files changed, 57 insertions(+), 98 deletions(-)
> 
> diff --git a/drivers/scsi/storvsc_drv.c b/drivers/scsi/storvsc_drv.c
> index b94d471..6f18e94 100644
> --- a/drivers/scsi/storvsc_drv.c
> +++ b/drivers/scsi/storvsc_drv.c
> @@ -721,29 +721,16 @@ static void cache_wwn(struct storvsc_device 
> *stor_device,
>   }
>  }
>  
> -static int storvsc_channel_init(struct hv_device *device, bool is_fc)
> +static int storvsc_execute_vstor_op(struct hv_device *device,
> + struct storvsc_cmd_request *request,
> + bool status_check)
>  {
> - struct storvsc_device *stor_device;
> - struct storvsc_cmd_request *request;
>   struct vstor_packet *vstor_packet;
> - int ret, t, i;
> - int max_chns;
> - bool process_sub_channels = false;
> -
> - stor_device = get_out_stor_device(device);
> - if (!stor_device)
> - return -ENODEV;
> + int ret, t;
>  
> - request = _device->init_request;
>   vstor_packet = >vstor_packet;
>  
> - /*
> -  * Now, initiate the vsc/vsp initialization protocol on the open
> -  * channel
> -  */
> - memset(request, 0, sizeof(struct storvsc_cmd_request));
>   init_completion(>wait_event);
> - vstor_packet->operation = VSTOR_OPERATION_BEGIN_INITIALIZATION;
>   vstor_packet->flags = REQUEST_COMPLETION_FLAG;
>  
>   ret = vmbus_sendpacket(device->channel, vstor_packet,
> @@ -753,27 +740,62 @@ static int storvsc_channel_init(struct hv_device 
> *device, bool is_fc)
>  VM_PKT_DATA_INBAND,
>  VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
>   if (ret != 0)
> - goto cleanup;
> + goto done;
>  
>   t = wait_for_completion_timeout(>wait_event, 5*HZ);
>   if (t == 0) {
>   ret = -ETIMEDOUT;
> - goto cleanup;
> + goto done;
>   }
>  
> + if (!status_check)
> + goto done;
> +
>   if (vstor_packet->operation != VSTOR_OPERATION_COMPLETE_IO ||
>   vstor_packet->status != 0) {
>   ret = -EINVAL;
> - goto cleanup;
> + goto done;
>   }
>  
> +done:
> + return ret;
> +}
> +
> +static int storvsc_channel_init(struct hv_device *device, bool is_fc)
> +{
> + struct storvsc_device *stor_device;
> + struct storvsc_cmd_request *request;
> + struct vstor_packet *vstor_packet;
> + int ret, i;
> + int max_chns;
> + bool process_sub_channels = false;
> +
> + stor_device = get_out_stor_device(device);
> + if (!stor_device)
> + return -ENODEV;
> +
> + request = _device->init_request;
> + vstor_packet = >vstor_packet;
> +
> + /*
> +  * Now, initiate the vsc/vsp initialization protocol on the open
> +  * channel
> +  */
> + memset(request, 0, sizeof(struct storvsc_cmd_request));
> + vstor_packet->operation = VSTOR_OPERATION_BEGIN_INITIALIZATION;
> + ret = storvsc_execute_vstor_op(device, request, true);
> + if (ret)
> + goto cleanup;
> +
> + /*
> +  * Query host supported protocol version.
> +  */
>  
>   for (i = 0; i < ARRAY_SIZE(vmstor_protocols); i++) {
>   /* reuse the packet for version range supported */
>   memset(vstor_packet, 0, sizeof(struct vstor_packet));
>   vstor_packet->operation =
>   VSTOR_OPERATION_QUERY_PROTOCOL_VERSION;
> - vstor_packet->flags = REQUEST_COMPLETION_FLAG;
>  
>   vstor_packet->version.major_minor =
>   vmstor_protocols[i].protocol_version;
> @@ -783,20 +805,9 @@ static int storvsc_channel_init(struct hv_device 
> *device, bool is_fc)
>*/
>   vstor_packet->version.revision = 0;
>  
> - ret = vmbus_sendpacket(device->channel, vstor_packet,
> -(sizeof(struct vstor_packet) -
> - vmscsi_size_delta),
> -(unsigned long)request,
> -VM_PKT_DATA_INBAND,
> -VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
> - if (ret != 0)
> - goto cleanup;
> -
> - t = wait_for_completion_timeout(>wait_event, 5*HZ);
> - if (t == 0) {
> - ret = -ETIMEDOUT;
> + ret = storvsc_execute_vstor_op(device, request, false);
> + if (ret)

RE: [PATCH 3/4] scsi: storvsc: Refactor the code in storvsc_channel_init()

[KY Srinivasan]

> -Original Message-
> From: Dan Carpenter [mailto:dan.carpen...@oracle.com]
> Sent: Friday, December 11, 2015 2:41 AM
> To: KY Srinivasan <k...@microsoft.com>
> Cc: gre...@linuxfoundation.org; linux-kernel@vger.kernel.org;
> de...@linuxdriverproject.org; oher...@suse.com;
> jbottom...@parallels.com; h...@infradead.org; linux-s...@vger.kernel.org;
> a...@canonical.com; vkuzn...@redhat.com; jasow...@redhat.com;
> martin.peter...@oracle.com
> Subject: Re: [PATCH 3/4] scsi: storvsc: Refactor the code in
> storvsc_channel_init()
> 
> On Thu, Dec 10, 2015 at 04:14:19PM -0800, K. Y. Srinivasan wrote:
> > @@ -753,27 +740,62 @@ static int storvsc_channel_init(struct hv_device
> *device, bool is_fc)
> >VM_PKT_DATA_INBAND,
> >
> VMBUS_DATA_PACKET_FLAG_COMPLETION_REQUESTED);
> > if (ret != 0)
> > -   goto cleanup;
> > +   goto done;
> >
> > t = wait_for_completion_timeout(>wait_event, 5*HZ);
> > if (t == 0) {
> > ret = -ETIMEDOUT;
> > -   goto cleanup;
> > +   goto done;
> > }
> >
> > +   if (!status_check)
> > +   goto done;
> 
> See?  This goto looks exactly the same as the earlier buggy goto but
> it's actually correct.  Meanwhile if you just used an explicit
> "return 0;" then it would be easy to understand.
> 
> I rant about this all the time but it's because it's bad deliberately.
> It's normal to have bugs, but this deliberate stuff really I can't
> understand it...
> 
> > +
> > if (vstor_packet->operation != VSTOR_OPERATION_COMPLETE_IO
> ||
> > vstor_packet->status != 0) {
> > ret = -EINVAL;
> > -   goto cleanup;
> > +   goto done;
> > }
> >
> > +done:
> > +   return ret;
> > +}
> > +
> > +static int storvsc_channel_init(struct hv_device *device, bool is_fc)
> > +{
> > +   struct storvsc_device *stor_device;
> > +   struct storvsc_cmd_request *request;
> > +   struct vstor_packet *vstor_packet;
> > +   int ret, i;
> > +   int max_chns;
> > +   bool process_sub_channels = false;
> > +
> > +   stor_device = get_out_stor_device(device);
> > +   if (!stor_device)
> > +   return -ENODEV;
> > +
> > +   request = _device->init_request;
> > +   vstor_packet = >vstor_packet;
> > +
> > +   /*
> > +* Now, initiate the vsc/vsp initialization protocol on the open
> > +* channel
> > +*/
> > +   memset(request, 0, sizeof(struct storvsc_cmd_request));
> > +   vstor_packet->operation =
> VSTOR_OPERATION_BEGIN_INITIALIZATION;
> > +   ret = storvsc_execute_vstor_op(device, request, true);
> > +   if (ret)
> > +   goto cleanup;
> 
> 10 lines earlier there is an explicit "return -ENODEV" so it's not as if
> writing explicit returns will kill you.

Thanks Dan; I will cleanup the code and resend.

Regards,

K. Y



--
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to majord...@vger.kernel.org
More majordomo info at  http://vger.kernel.org/majordomo-info.html
Please read the FAQ at  http://www.tux.org/lkml/