Skip to contentSkip to conversation

net/dpaa2: fix build error about timesync functions

PATCH Superseded Chenxu Di 6 participants 17 messages Sep 15, 2020 – Oct 9, 2020

Subsystems (2)
lore
Share & tools
lore

Chenxu Di

author
#0 lore raw

PATCH 1/1 net/dpaa2: fix build error about timesync functions

When the build option has '-DRTE_LIBRTE_IEEE1588=1', the announce of timesync functions will be build. However the dpdk_conf doesn't hav RTE_LIBRTE_IEEE1588 so that the file dpaa2_ptp.c will not be build. It cause the build error.
This patch fixes it by adding set for dpdk_conf.

Fixes 184c39d16568 ("net/dpaa2: add DPRTC sub-module")
---
 drivers/net/dpaa2/meson.build | 4 ++++
 1 file changed, 4 insertions(+)
drivers/net/dpaa2/meson.build +4 -0
index 6dd0eb274..d9aadfdae 100644
@@ -17,6 +17,10 @@ sources = files('base/dpaa2_hw_dpni.c',
 		'mc/dpdmux.c',
 		'mc/dpni.c')
 
+if '-DRTE_LIBRTE_IEEE1588=1' in get_option('c_args')
+	dpdk_conf.set('RTE_LIBRTE_IEEE1588', 1)
+endif
+
 if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
 	sources += files('mc/dprtc.c')
 	sources += files('dpaa2_ptp.c')
Signature · 1 line
2.17.1
Tested-by on v1 1/1
#2 lore raw
In reply to Chenxu Di

-----Original Message-----
From: dev [mailto:[email protected]] On Behalf Of Chenxu Di
Sent: Tuesday, September 15, 2020 10:41 AM
To: [email protected]
Cc: [email protected]; [email protected]; Di, ChenxuX <[email protected]>; [email protected]
Subject: [dpdk-dev] [PATCH] net/dpaa2: fix build error about timesync functions

When the build option has '-DRTE_LIBRTE_IEEE1588=1', the announce of timesync functions will be build. However the dpdk_conf doesn't hav RTE_LIBRTE_IEEE1588 so that the file dpaa2_ptp.c will not be build. It cause the build error.
This patch fixes it by adding set for dpdk_conf.

Fixes 184c39d16568 ("net/dpaa2: add DPRTC sub-module")
---
 drivers/net/dpaa2/meson.build | 4 ++++
 1 file changed, 4 insertions(+)
drivers/net/dpaa2/meson.build +3 -0
@@ -17,6 +17,10 @@ sources = files('base/dpaa2_hw_dpni.c',
 		'mc/dpdmux.c',
 		'mc/dpni.c')
 
+if '-DRTE_LIBRTE_IEEE1588=1' in get_option('c_args')
+	dpdk_conf.set('RTE_LIBRTE_IEEE1588', 1) endif
+
 if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
 	sources += files('mc/dprtc.c')
 	sources += files('dpaa2_ptp.c')

--
2.17.1

#3 lore raw
In reply to Chenxu Di
Quoted 20 lines · Chenxu Di

When the build option has '-DRTE_LIBRTE_IEEE1588=1', the announce of timesync functions will be build. However the dpdk_conf doesn't hav RTE_LIBRTE_IEEE1588 so that the file dpaa2_ptp.c will not be build. It cause the build error.
This patch fixes it by adding set for dpdk_conf.

Fixes 184c39d16568 ("net/dpaa2: add DPRTC sub-module")
---
  drivers/net/dpaa2/meson.build | 4 ++++
  1 file changed, 4 insertions(+)
drivers/net/dpaa2/meson.build +0 -0
index 6dd0eb274..d9aadfdae 100644

drivers/net/dpaa2/meson.build sources = files('base/dpaa2_hw_dpni.c', · line 20

Quoted 4 lines
drivers/net/dpaa2/meson.build +1 -0
@@ -17,6 +17,10 @@ sources = files('base/dpaa2_hw_dpni.c',
  		'mc/dpdmux.c',
  		'mc/dpni.c')
 
+if '-DRTE_LIBRTE_IEEE1588=1' in get_option('c_args')

The "RTE_LIBRTE_IEEE1588=1" can fail,
all places looking for "#ifdef RTE_LIBRTE_IEEE1588", so a "-Dc_args=-DRTE_LIBRTE_IEEE1588" is more likely, but why not "-Dc_args=-DRTE_LIBRTE_IEEE1588=666"

drivers/net/dpaa2/meson.build sources = files('base/dpaa2_hw_dpni.c', · line 26

Quoted 7 lines
drivers/net/dpaa2/meson.build +3 -0
+	dpdk_conf.set('RTE_LIBRTE_IEEE1588', 1)
+endif
+
  if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
  	sources += files('mc/dprtc.c')
  	sources += files('dpaa2_ptp.c')

Can't we just remove the conditional build:

-if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
- sources += files('mc/dprtc.c')
- sources += files('dpaa2_ptp.c')
-endif
+sources += files('mc/dprtc.c')
+sources += files('dpaa2_ptp.c')

#4 lore raw

Hi,

Quoted 33 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Wednesday, September 16, 2020 11:29 PM
To: Di, ChenxuX <[email protected]>; [email protected]
Cc: [email protected]; [email protected]; [email protected]; Richardson, Bruce <[email protected]>
Subject: Re: [dpdk-stable] [PATCH] net/dpaa2: fix build error about timesync functions

Quoted 23 lines · Chenxu Di

When the build option has '-DRTE_LIBRTE_IEEE1588=1', the announce of timesync functions will be build.
However the dpdk_conf doesn't hav RTE_LIBRTE_IEEE1588 so that the file dpaa2_ptp.c will not be build.
It cause the build error.
This patch fixes it by adding set for dpdk_conf.

Fixes 184c39d16568 ("net/dpaa2: add DPRTC sub-module")
---
  drivers/net/dpaa2/meson.build | 4 ++++
  1 file changed, 4 insertions(+)
(quoted diff) +0 -0

b/drivers/net/dpaa2/meson.build index 6dd0eb274..d9aadfdae 100644

drivers/net/dpaa2/meson.build +0 -0
@@ -17,6 +17,10 @@ sources = files('base/dpaa2_hw_dpni.c',
  		'mc/dpdmux.c',
  		'mc/dpni.c')
 
Quoted 7 lines

drivers/net/dpaa2/meson.build sources = files('base/dpaa2_hw_dpni.c', · line 20

Quoted 1 line
drivers/net/dpaa2/meson.build +1 -0
+if '-DRTE_LIBRTE_IEEE1588=1' in get_option('c_args')

The "RTE_LIBRTE_IEEE1588=1" can fail,
all places looking for "#ifdef RTE_LIBRTE_IEEE1588", so a "-Dc_args=- DRTE_LIBRTE_IEEE1588" is more likely, but why not "-Dc_args=- DRTE_LIBRTE_IEEE1588=666"

Yes, I will change it

Quoted 9 lines

drivers/net/dpaa2/meson.build sources = files('base/dpaa2_hw_dpni.c', · line 26

Quoted 6 lines
drivers/net/dpaa2/meson.build +2 -0
+	dpdk_conf.set('RTE_LIBRTE_IEEE1588', 1) endif
+
  if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
  	sources += files('mc/dprtc.c')
  	sources += files('dpaa2_ptp.c')

Can't we just remove the conditional build:

Quoted 6 lines

-if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
- sources += files('mc/dprtc.c')
- sources += files('dpaa2_ptp.c')
-endif
+sources += files('mc/dprtc.c')
+sources += files('dpaa2_ptp.c')

The announce of timesync functions are in the #define DRTE_LIBRTE_IEEE1588
While the define of the functions are in the file 'dpaa2_ptp.c'.
So they should be both build or not build by whether the build option -DRTE_LIBRTE_IEEE1588=1 or not. So it seems not a good idea that remove the conditional.

#5 lore raw
Quoted 61 lines · ChenxuX

Hi,

Quoted 33 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Wednesday, September 16, 2020 11:29 PM
To: Di, ChenxuX <[email protected]>; [email protected]
Cc: [email protected]; [email protected]; [email protected]; Richardson, Bruce <[email protected]>
Subject: Re: [dpdk-stable] [PATCH] net/dpaa2: fix build error about timesync functions

Quoted 23 lines · Chenxu Di

When the build option has '-DRTE_LIBRTE_IEEE1588=1', the announce of timesync functions will be build.
However the dpdk_conf doesn't hav RTE_LIBRTE_IEEE1588 so that the file dpaa2_ptp.c will not be build.
It cause the build error.
This patch fixes it by adding set for dpdk_conf.

Fixes 184c39d16568 ("net/dpaa2: add DPRTC sub-module")
---
   drivers/net/dpaa2/meson.build | 4 ++++
   1 file changed, 4 insertions(+)
(quoted diff) +0 -0

b/drivers/net/dpaa2/meson.build index 6dd0eb274..d9aadfdae 100644

drivers/net/dpaa2/meson.build +0 -0
@@ -17,6 +17,10 @@ sources = files('base/dpaa2_hw_dpni.c',
   		'mc/dpdmux.c',
   		'mc/dpni.c')
 
Quoted 7 lines

drivers/net/dpaa2/meson.build sources = files('base/dpaa2_hw_dpni.c', · line 20

Quoted 1 line
drivers/net/dpaa2/meson.build +1 -0
+if '-DRTE_LIBRTE_IEEE1588=1' in get_option('c_args')

The "RTE_LIBRTE_IEEE1588=1" can fail,
all places looking for "#ifdef RTE_LIBRTE_IEEE1588", so a "-Dc_args=- DRTE_LIBRTE_IEEE1588" is more likely, but why not "-Dc_args=- DRTE_LIBRTE_IEEE1588=666"

Yes, I will change it

Quoted 9 lines

drivers/net/dpaa2/meson.build sources = files('base/dpaa2_hw_dpni.c', · line 26

Quoted 6 lines
drivers/net/dpaa2/meson.build +2 -0
+	dpdk_conf.set('RTE_LIBRTE_IEEE1588', 1) endif
+
   if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
   	sources += files('mc/dprtc.c')
   	sources += files('dpaa2_ptp.c')

Can't we just remove the conditional build:

Quoted 6 lines
   -if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
   -       sources += files('mc/dprtc.c')
   -       sources += files('dpaa2_ptp.c')
   -endif
   +sources += files('mc/dprtc.c')
   +sources += files('dpaa2_ptp.c')
Quoted 5 lines

The announce of timesync functions are in the #define DRTE_LIBRTE_IEEE1588
While the define of the functions are in the file 'dpaa2_ptp.c'.
So they should be both build or not build by whether the build option -DRTE_LIBRTE_IEEE1588=1 or not. So it seems not a good idea that remove the conditional.

timesyncs_* dev_ops functions defined but not used is not big problem, only can increase the library size.

I believe more concern is on:
'RTE_PMD_REGISTER_DPAA2_OBJECT(dprtc, rte_dpaa2_dprtc_obj);'
which looks like register function to run in constructor when 'dpaa2_ptp.c' is compiled, but not sure affect of it. It can be possible to wrap that call with 'RTE_LIBRTE_IEEE1588' ifdef, that should work.

It is preferred to remove the compile time flag instead of finding ways to make it work.

Let's wait for the dpaa2 maintainers' response, perhaps they can come with a smart way to remove the compile time flag.

#6 lore raw
Quoted 71 lines · Ferruh Yigit
Quoted 61 lines · ChenxuX

Hi,

Quoted 33 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Wednesday, September 16, 2020 11:29 PM
To: Di, ChenxuX <[email protected]>; [email protected]
Cc: [email protected]; [email protected]; [email protected]; Richardson, Bruce <[email protected]>
Subject: Re: [dpdk-stable] [PATCH] net/dpaa2: fix build error about timesync functions

Quoted 23 lines · Chenxu Di

When the build option has '-DRTE_LIBRTE_IEEE1588=1', the announce of timesync functions will be build.
However the dpdk_conf doesn't hav RTE_LIBRTE_IEEE1588 so that the file dpaa2_ptp.c will not be build.
It cause the build error.
This patch fixes it by adding set for dpdk_conf.

Fixes 184c39d16568 ("net/dpaa2: add DPRTC sub-module")

---
   drivers/net/dpaa2/meson.build | 4 ++++

   1 file changed, 4 insertions(+)
(quoted diff) +0 -0

b/drivers/net/dpaa2/meson.build index 6dd0eb274..d9aadfdae 100644

drivers/net/dpaa2/meson.build +0 -0
@@ -17,6 +17,10 @@ sources = files('base/dpaa2_hw_dpni.c',
            'mc/dpdmux.c',
            'mc/dpni.c')
 
Quoted 7 lines

drivers/net/dpaa2/meson.build sources = files('base/dpaa2_hw_dpni.c', · line 20

Quoted 1 line
drivers/net/dpaa2/meson.build +1 -0
+if '-DRTE_LIBRTE_IEEE1588=1' in get_option('c_args')

The "RTE_LIBRTE_IEEE1588=1" can fail,
all places looking for "#ifdef RTE_LIBRTE_IEEE1588", so a "-Dc_args=- DRTE_LIBRTE_IEEE1588" is more likely, but why not "-Dc_args=- DRTE_LIBRTE_IEEE1588=666"

Yes, I will change it

Quoted 7 lines

drivers/net/dpaa2/meson.build sources = files('base/dpaa2_hw_dpni.c', · line 26

Quoted 6 lines
drivers/net/dpaa2/meson.build +2 -0
+ dpdk_conf.set('RTE_LIBRTE_IEEE1588', 1) endif
+
    if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
        sources += files('mc/dprtc.c')
        sources += files('dpaa2_ptp.c')
Quoted 8 lines

Can't we just remove the conditional build:

   -if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
   -       sources += files('mc/dprtc.c')
   -       sources += files('dpaa2_ptp.c')
   -endif
   +sources += files('mc/dprtc.c')
   +sources += files('dpaa2_ptp.c')

Quoted 6 lines

The announce of timesync functions are in the #define DRTE_LIBRTE_IEEE1588 While the define of the functions are in the file 'dpaa2_ptp.c'. So they should be both build or not build by whether the build option -DRTE_LIBRTE_IEEE1588=1 or not.
So it seems not a good idea that remove the conditional.

timesyncs_* dev_ops functions defined but not used is not big problem, only can increase the library size.

Quoted 8 lines

I believe more concern is on:
'RTE_PMD_REGISTER_DPAA2_OBJECT(dprtc, rte_dpaa2_dprtc_obj);'
which looks like register function to run in constructor when 'dpaa2_ptp.c' is compiled, but not sure affect of it. It can be possible to wrap that call with 'RTE_LIBRTE_IEEE1588' ifdef, that should work.

It is preferred to remove the compile time flag instead of finding ways to make it work.

Let's wait for the dpaa2 maintainers' response, perhaps they can come with a smart way to remove the compile time flag.

Hi Ferruh,

   enabling  IEEE1588 causes some performance drop in the dpaa2 performance. That is the reason, we have kept this code in compile time flag.

However, we will work in future to make it run-time configurable but that will require some code restructuring and it will be a moderate size work.

Regards,

Hemant

#7 lore raw
Quoted 82 lines · Hemant Agrawal
Quoted 71 lines · Ferruh Yigit
Quoted 61 lines · ChenxuX

Hi,

Quoted 33 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Wednesday, September 16, 2020 11:29 PM
To: Di, ChenxuX <[email protected]>; [email protected]
Cc: [email protected]; [email protected]; [email protected]; Richardson, Bruce <[email protected]>
Subject: Re: [dpdk-stable] [PATCH] net/dpaa2: fix build error about timesync functions

Quoted 23 lines · Chenxu Di

When the build option has '-DRTE_LIBRTE_IEEE1588=1', the announce of timesync functions will be build.
However the dpdk_conf doesn't hav RTE_LIBRTE_IEEE1588 so that the file dpaa2_ptp.c will not be build.
It cause the build error.
This patch fixes it by adding set for dpdk_conf.

Fixes 184c39d16568 ("net/dpaa2: add DPRTC sub-module")

---
   drivers/net/dpaa2/meson.build | 4 ++++

   1 file changed, 4 insertions(+)
(quoted diff) +0 -0

b/drivers/net/dpaa2/meson.build index 6dd0eb274..d9aadfdae 100644

drivers/net/dpaa2/meson.build +0 -0
@@ -17,6 +17,10 @@ sources = files('base/dpaa2_hw_dpni.c',
            'mc/dpdmux.c',
            'mc/dpni.c')
 
Quoted 7 lines

drivers/net/dpaa2/meson.build sources = files('base/dpaa2_hw_dpni.c', · line 20

Quoted 1 line
drivers/net/dpaa2/meson.build +1 -0
+if '-DRTE_LIBRTE_IEEE1588=1' in get_option('c_args')

The "RTE_LIBRTE_IEEE1588=1" can fail,
all places looking for "#ifdef RTE_LIBRTE_IEEE1588", so a "-Dc_args=- DRTE_LIBRTE_IEEE1588" is more likely, but why not "-Dc_args=- DRTE_LIBRTE_IEEE1588=666"

Yes, I will change it

Quoted 7 lines

drivers/net/dpaa2/meson.build sources = files('base/dpaa2_hw_dpni.c', · line 26

Quoted 6 lines
drivers/net/dpaa2/meson.build +2 -0
+ dpdk_conf.set('RTE_LIBRTE_IEEE1588', 1) endif
+
    if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
        sources += files('mc/dprtc.c')
        sources += files('dpaa2_ptp.c')
Quoted 8 lines

Can't we just remove the conditional build:

   -if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
   -       sources += files('mc/dprtc.c')
   -       sources += files('dpaa2_ptp.c')
   -endif
   +sources += files('mc/dprtc.c')
   +sources += files('dpaa2_ptp.c')

Quoted 6 lines

The announce of timesync functions are in the #define DRTE_LIBRTE_IEEE1588 While the define of the functions are in the file 'dpaa2_ptp.c'. So they should be both build or not build by whether the build option -DRTE_LIBRTE_IEEE1588=1 or not.
So it seems not a good idea that remove the conditional.

timesyncs_* dev_ops functions defined but not used is not big problem, only can increase the library size.

Quoted 8 lines

I believe more concern is on:
'RTE_PMD_REGISTER_DPAA2_OBJECT(dprtc, rte_dpaa2_dprtc_obj);'
which looks like register function to run in constructor when 'dpaa2_ptp.c' is compiled, but not sure affect of it. It can be possible to wrap that call with 'RTE_LIBRTE_IEEE1588' ifdef, that should work.

It is preferred to remove the compile time flag instead of finding ways to make it work.

Let's wait for the dpaa2 maintainers' response, perhaps they can come with a smart way to remove the compile time flag.

Quoted 8 lines

Hi Ferruh,

   enabling  IEEE1588 causes some performance drop in the dpaa2
performance. That is the reason, we have kept this code in compile time flag.

However, we will work in future to make it run-time configurable but that will require some code restructuring and it will be a moderate size work.

OK, thanks Hemant.

At least can it be possible to remove it from the build files, what do you think about wrapping those two files (or their relevant parts) with 'RTE_LIBRTE_IEEE1588' ifdef and remove the checks from meson file?

#8 lore raw
Quoted 87 lines · Ferruh Yigit
Quoted 86 lines · Hemant Agrawal
Quoted 70 lines · Ferruh Yigit
Quoted 61 lines · ChenxuX

Hi,

Quoted 33 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Wednesday, September 16, 2020 11:29 PM
To: Di, ChenxuX <[email protected]>; [email protected]
Cc: [email protected]; [email protected]; [email protected]; Richardson, Bruce <[email protected]>
Subject: Re: [dpdk-stable] [PATCH] net/dpaa2: fix build error about timesync functions

Quoted 23 lines · Chenxu Di

When the build option has '-DRTE_LIBRTE_IEEE1588=1', the announce of timesync functions will be build.
However the dpdk_conf doesn't hav RTE_LIBRTE_IEEE1588 so that the file dpaa2_ptp.c will not be build.
It cause the build error.
This patch fixes it by adding set for dpdk_conf.

Fixes 184c39d16568 ("net/dpaa2: add DPRTC sub-module")

---
   drivers/net/dpaa2/meson.build | 4 ++++

   1 file changed, 4 insertions(+)
(quoted diff) +0 -0

b/drivers/net/dpaa2/meson.build index 6dd0eb274..d9aadfdae 100644

drivers/net/dpaa2/meson.build +0 -0
@@ -17,6 +17,10 @@ sources = files('base/dpaa2_hw_dpni.c',
            'mc/dpdmux.c',
            'mc/dpni.c')
 
Quoted 7 lines

drivers/net/dpaa2/meson.build sources = files('base/dpaa2_hw_dpni.c', · line 20

Quoted 1 line
drivers/net/dpaa2/meson.build +1 -0
+if '-DRTE_LIBRTE_IEEE1588=1' in get_option('c_args')

The "RTE_LIBRTE_IEEE1588=1" can fail,
all places looking for "#ifdef RTE_LIBRTE_IEEE1588", so a "-Dc_args=- DRTE_LIBRTE_IEEE1588" is more likely, but why not "-Dc_args=- DRTE_LIBRTE_IEEE1588=666"

Yes, I will change it

Quoted 7 lines

drivers/net/dpaa2/meson.build sources = files('base/dpaa2_hw_dpni.c', · line 26

Quoted 6 lines
drivers/net/dpaa2/meson.build +2 -0
+ dpdk_conf.set('RTE_LIBRTE_IEEE1588', 1) endif
+
    if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
        sources += files('mc/dprtc.c')
        sources += files('dpaa2_ptp.c')
Quoted 8 lines

Can't we just remove the conditional build:

   -if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
   -       sources += files('mc/dprtc.c')
   -       sources += files('dpaa2_ptp.c')
   -endif
   +sources += files('mc/dprtc.c')
   +sources += files('dpaa2_ptp.c')

Quoted 5 lines

The announce of timesync functions are in the #define DRTE_LIBRTE_IEEE1588
While the define of the functions are in the file 'dpaa2_ptp.c'.
So they should be both build or not build by whether the build option -DRTE_LIBRTE_IEEE1588=1 or not. So it seems not a good idea that remove the conditional.

timesyncs_* dev_ops functions defined but not used is not big problem, only can increase the library size.

Quoted 8 lines

I believe more concern is on:
'RTE_PMD_REGISTER_DPAA2_OBJECT(dprtc, rte_dpaa2_dprtc_obj);'
which looks like register function to run in constructor when 'dpaa2_ptp.c' is compiled, but not sure affect of it. It can be possible to wrap that call with 'RTE_LIBRTE_IEEE1588' ifdef, that should work.

It is preferred to remove the compile time flag instead of finding ways to make it work.

Let's wait for the dpaa2 maintainers' response, perhaps they can come with a smart way to remove the compile time flag.

Hi Ferruh,

    enabling  IEEE1588 causes some performance drop in the dpaa2 performance. That is the reason, we have kept this code in compile time flag.

Quoted 7 lines
Quoted 2 lines

However, we will work in future to make it run-time configurable but that will require some code restructuring and it will be a moderate size work.

OK, thanks Hemant.

At least can it be possible to remove it from the build files, what do you think about wrapping those two files (or their relevant parts) with 'RTE_LIBRTE_IEEE1588' ifdef and remove the checks from meson file?

ok.

#9 lore raw
Quoted 89 lines · Hemant Agrawal
Quoted 88 lines · Ferruh Yigit
Quoted 87 lines · Hemant Agrawal
Quoted 71 lines · Ferruh Yigit
Quoted 61 lines · ChenxuX

Hi,

Quoted 33 lines

-----Original Message-----
From: Ferruh Yigit <[email protected]>
Sent: Wednesday, September 16, 2020 11:29 PM
To: Di, ChenxuX <[email protected]>; [email protected]
Cc: [email protected]; [email protected]; [email protected]; Richardson, Bruce <[email protected]>
Subject: Re: [dpdk-stable] [PATCH] net/dpaa2: fix build error about timesync functions

Quoted 23 lines · Chenxu Di

When the build option has '-DRTE_LIBRTE_IEEE1588=1', the announce of timesync functions will be build.
However the dpdk_conf doesn't hav RTE_LIBRTE_IEEE1588 so that the file dpaa2_ptp.c will not be build.
It cause the build error.
This patch fixes it by adding set for dpdk_conf.

Fixes 184c39d16568 ("net/dpaa2: add DPRTC sub-module")

---
   drivers/net/dpaa2/meson.build | 4 ++++

   1 file changed, 4 insertions(+)
(quoted diff) +0 -0

b/drivers/net/dpaa2/meson.build index 6dd0eb274..d9aadfdae 100644

drivers/net/dpaa2/meson.build +0 -0
@@ -17,6 +17,10 @@ sources = files('base/dpaa2_hw_dpni.c',
            'mc/dpdmux.c',
            'mc/dpni.c')
 
Quoted 7 lines

drivers/net/dpaa2/meson.build sources = files('base/dpaa2_hw_dpni.c', · line 20

Quoted 1 line
drivers/net/dpaa2/meson.build +1 -0
+if '-DRTE_LIBRTE_IEEE1588=1' in get_option('c_args')

The "RTE_LIBRTE_IEEE1588=1" can fail,
all places looking for "#ifdef RTE_LIBRTE_IEEE1588", so a "-Dc_args=- DRTE_LIBRTE_IEEE1588" is more likely, but why not "-Dc_args=- DRTE_LIBRTE_IEEE1588=666"

Yes, I will change it

Quoted 7 lines

drivers/net/dpaa2/meson.build sources = files('base/dpaa2_hw_dpni.c', · line 25

Quoted 6 lines
drivers/net/dpaa2/meson.build +2 -0
+ dpdk_conf.set('RTE_LIBRTE_IEEE1588', 1) endif
+
    if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
        sources += files('mc/dprtc.c')
        sources += files('dpaa2_ptp.c')
Quoted 8 lines

Can't we just remove the conditional build:

   -if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
   -       sources += files('mc/dprtc.c')
   -       sources += files('dpaa2_ptp.c')
   -endif
   +sources += files('mc/dprtc.c')
   +sources += files('dpaa2_ptp.c')

Quoted 6 lines

The announce of timesync functions are in the #define DRTE_LIBRTE_IEEE1588 While the define of the functions are in the file 'dpaa2_ptp.c'. So they should be both build or not build by whether the build option -DRTE_LIBRTE_IEEE1588=1 or not.
So it seems not a good idea that remove the conditional.

timesyncs_* dev_ops functions defined but not used is not big problem, only can increase the library size.

Quoted 8 lines

I believe more concern is on:
'RTE_PMD_REGISTER_DPAA2_OBJECT(dprtc, rte_dpaa2_dprtc_obj);'
which looks like register function to run in constructor when 'dpaa2_ptp.c' is compiled, but not sure affect of it. It can be possible to wrap that call with 'RTE_LIBRTE_IEEE1588' ifdef, that should work.

It is preferred to remove the compile time flag instead of finding ways to make it work.

Let's wait for the dpaa2 maintainers' response, perhaps they can come with a smart way to remove the compile time flag.

Hi Ferruh,

    enabling  IEEE1588 causes some performance drop in the dpaa2 performance. That is the reason, we have kept this code in compile time flag.

Quoted 8 lines
Quoted 7 lines
Quoted 2 lines

However, we will work in future to make it run-time configurable but that will require some code restructuring and it will be a moderate size work.

OK, thanks Hemant.

At least can it be possible to remove it from the build files, what do you think about wrapping those two files (or their relevant parts) with 'RTE_LIBRTE_IEEE1588' ifdef and remove the checks from meson file?

ok.

Hi Hemant, Chenxu,

Can you send a new version of the patch as discussed above?

Thanks,
ferruh

author
#10 lore raw

PATCH v2 1/1 net/dpaa2: fix build error about timesync functions

In reply to Chenxu Di

When the build option has '-DRTE_LIBRTE_IEEE1588=1', the announce of timesync functions will be build. However the dpdk_conf doesn't hav RTE_LIBRTE_IEEE1588 so that the file dpaa2_ptp.c will not be build. It cause the build error.
This patch fixes it by adding set for dpdk_conf.

Fixes 184c39d16568 ("net/dpaa2: add DPRTC sub-module")
---
 drivers/net/dpaa2/meson.build | 6 ++++++
 1 file changed, 6 insertions(+)
drivers/net/dpaa2/meson.build +6 -0
index 6dd0eb274..43d625dc5 100644
@@ -17,6 +17,12 @@ sources = files('base/dpaa2_hw_dpni.c',
 		'mc/dpdmux.c',
 		'mc/dpni.c')
 
+foreach each:get_option('c_args')
+	if each.contains('-DRTE_LIBRTE_IEEE1588')
+		dpdk_conf.set('RTE_LIBRTE_IEEE1588', 1)
+	endif
+endforeach
+
 if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
 	sources += files('mc/dprtc.c')
 	sources += files('dpaa2_ptp.c')
Signature · 1 line
2.17.1
Acked-by on v2 1/1
#11 lore raw
Quoted 32 lines · Chenxu Di

When the build option has '-DRTE_LIBRTE_IEEE1588=1', the announce of timesync functions will be build. However the dpdk_conf doesn't hav RTE_LIBRTE_IEEE1588 so that the file dpaa2_ptp.c will not be build. It cause the build error.
This patch fixes it by adding set for dpdk_conf.

Fixes 184c39d16568 ("net/dpaa2: add DPRTC sub-module")
---
  drivers/net/dpaa2/meson.build | 6 ++++++
  1 file changed, 6 insertions(+)
drivers/net/dpaa2/meson.build +6 -0
index 6dd0eb274..43d625dc5 100644
@@ -17,6 +17,12 @@ sources = files('base/dpaa2_hw_dpni.c',
  		'mc/dpdmux.c',
  		'mc/dpni.c')
 
+foreach each:get_option('c_args')
+	if each.contains('-DRTE_LIBRTE_IEEE1588')
+		dpdk_conf.set('RTE_LIBRTE_IEEE1588', 1)
+	endif
+endforeach
+
  if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
  	sources += files('mc/dprtc.c')
  	sources += files('dpaa2_ptp.c')
#12 lore raw

PATCH v3 1/1 net/dpaa2: fix build error about timesync functions

In reply to Chenxu Di

From: Chenxu Di <[email protected]>

Some timesync related source files are built only when 'RTE_LIBRTE_IEEE1588' config is set, which is missing in meson.

Building with '-DRTE_LIBRTE_IEEE1588' cflag will enable timesync functionality in the files that are build, but it won't build files that require 'RTE_LIBRTE_IEEE1588' config.
This causes the build error.

Fixing by removing config check in meson file and build all .c files by default, but wrap relevant code part with 'RTE_LIBRTE_IEEE1588' macro.

Also removing 'RTE_LIBRTE_IEEE1588' ifdef around some fields of the data structures, to not require finer grained macro wraps on the functions using them.
Since the registration of the function disabled with macro check, having functions compiled shouldn't affect the functionality.

Fixes 184c39d16568 ("net/dpaa2: add DPRTC sub-module")
---
 drivers/net/dpaa2/dpaa2_ethdev.h | 4 ----
 drivers/net/dpaa2/dpaa2_ptp.c    | 2 ++
 drivers/net/dpaa2/meson.build    | 7 ++-----
 3 files changed, 4 insertions(+), 9 deletions(-)
drivers/net/dpaa2/dpaa2_ethdev.h +0 -4
index e577355d57..94cf253827 100644
@@ -160,7 +160,6 @@ struct dpaa2_dev_priv {
 	uint16_t ss_offset;
 	uint64_t ss_iova;
 	uint64_t ss_param_iova;
-#if defined(RTE_LIBRTE_IEEE1588)
 	/*stores timestamp of last received packet on dev*/
 	uint64_t rx_timestamp;
 	/*stores timestamp of last received tx confirmation packet on dev*/
@@ -169,7 +168,6 @@ struct dpaa2_dev_priv {
 	 * it corresponds to last packet transmitted
 	 */
 	struct dpaa2_queue *next_tx_conf_queue;
-#endif
 
 	struct rte_eth_dev *eth_dev; /**< Pointer back to holding ethdev */
 
@@ -227,7 +225,6 @@ void dpaa2_dev_free_eqresp_buf(uint16_t eqresp_ci);
 void dpaa2_flow_clean(struct rte_eth_dev *dev);
 uint16_t dpaa2_dev_tx_conf(void *queue)  __rte_unused;
 
-#if defined(RTE_LIBRTE_IEEE1588)
 int dpaa2_timesync_enable(struct rte_eth_dev *dev);
 int dpaa2_timesync_disable(struct rte_eth_dev *dev);
 int dpaa2_timesync_read_time(struct rte_eth_dev *dev,
@@ -240,5 +237,4 @@ int dpaa2_timesync_read_rx_timestamp(struct rte_eth_dev *dev,
 						uint32_t flags __rte_unused);
 int dpaa2_timesync_read_tx_timestamp(struct rte_eth_dev *dev,
 					  struct timespec *timestamp);
-#endif
 #endif /* _DPAA2_ETHDEV_H */
drivers/net/dpaa2/dpaa2_ptp.c +2 -0
index f58eedb314..899dd5d442 100644
@@ -129,6 +129,7 @@ int dpaa2_timesync_read_rx_timestamp(struct rte_eth_dev *dev,
 	return 0;
 }
 
+#if defined(RTE_LIBRTE_IEEE1588)
 static int
 dpaa2_create_dprtc_device(int vdev_fd __rte_unused,
 			   struct vfio_device_info *obj_info __rte_unused,
@@ -179,3 +180,4 @@ static struct rte_dpaa2_object rte_dpaa2_dprtc_obj = {
 };
 
 RTE_PMD_REGISTER_DPAA2_OBJECT(dprtc, rte_dpaa2_dprtc_obj);
+#endif
drivers/net/dpaa2/meson.build +2 -5
index 6dd0eb274c..4312aa73f7 100644
@@ -13,15 +13,12 @@ sources = files('base/dpaa2_hw_dpni.c',
 		'dpaa2_flow.c',
 		'dpaa2_rxtx.c',
 		'dpaa2_sparser.c',
+		'dpaa2_ptp.c',
+		'mc/dprtc.c',
 		'mc/dpkg.c',
 		'mc/dpdmux.c',
 		'mc/dpni.c')
 
-if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
-	sources += files('mc/dprtc.c')
-	sources += files('dpaa2_ptp.c')
-endif
-
 includes += include_directories('base', 'mc')
 
 install_headers('rte_pmd_dpaa2.h')
Signature · 1 line
2.26.2
#13 lore raw

Following change should also be handled as suggested by Ferruh in earlier.

   diff --git a/drivers/net/dpaa2/dpaa2_ptp.c
   b/drivers/net/dpaa2/dpaa2_ptp.c
   index 9f755e84bf..0e44d4c6bf 100644
   --- a/drivers/net/dpaa2/dpaa2_ptp.c
   +++ b/drivers/net/dpaa2/dpaa2_ptp.c
   @@ -178,4 +178,6 @@ static struct rte_dpaa2_object
   rte_dpaa2_dprtc_obj = {
            .create = dpaa2_create_dprtc_device,
     };

   +#if defined(RTE_LIBRTE_IEEE1588)
     RTE_PMD_REGISTER_DPAA2_OBJECT(dprtc, rte_dpaa2_dprtc_obj);
   +#endif

regards,
Sachin

Quoted 103 lines · Ferruh Yigit

From: Chenxu Di <[email protected]>

Some timesync related source files are built only when 'RTE_LIBRTE_IEEE1588' config is set, which is missing in meson.

Building with '-DRTE_LIBRTE_IEEE1588' cflag will enable timesync functionality in the files that are build, but it won't build files that require 'RTE_LIBRTE_IEEE1588' config.
This causes the build error.

Fixing by removing config check in meson file and build all .c files by default, but wrap relevant code part with 'RTE_LIBRTE_IEEE1588' macro.

Also removing 'RTE_LIBRTE_IEEE1588' ifdef around some fields of the data structures, to not require finer grained macro wraps on the functions using them.
Since the registration of the function disabled with macro check, having functions compiled shouldn't affect the functionality.

Fixes 184c39d16568 ("net/dpaa2: add DPRTC sub-module")
---
  drivers/net/dpaa2/dpaa2_ethdev.h | 4 ----
  drivers/net/dpaa2/dpaa2_ptp.c    | 2 ++
  drivers/net/dpaa2/meson.build    | 7 ++-----
  3 files changed, 4 insertions(+), 9 deletions(-)
drivers/net/dpaa2/dpaa2_ethdev.h +0 -4
index e577355d57..94cf253827 100644
@@ -160,7 +160,6 @@ struct dpaa2_dev_priv {
  	uint16_t ss_offset;
  	uint64_t ss_iova;
  	uint64_t ss_param_iova;
-#if defined(RTE_LIBRTE_IEEE1588)
  	/*stores timestamp of last received packet on dev*/
  	uint64_t rx_timestamp;
  	/*stores timestamp of last received tx confirmation packet on dev*/
@@ -169,7 +168,6 @@ struct dpaa2_dev_priv {
  	 * it corresponds to last packet transmitted
  	 */
  	struct dpaa2_queue *next_tx_conf_queue;
-#endif
 
  	struct rte_eth_dev *eth_dev; /**< Pointer back to holding ethdev */
 
@@ -227,7 +225,6 @@ void dpaa2_dev_free_eqresp_buf(uint16_t eqresp_ci);
  void dpaa2_flow_clean(struct rte_eth_dev *dev);
  uint16_t dpaa2_dev_tx_conf(void *queue)  __rte_unused;
 
-#if defined(RTE_LIBRTE_IEEE1588)
  int dpaa2_timesync_enable(struct rte_eth_dev *dev);
  int dpaa2_timesync_disable(struct rte_eth_dev *dev);
  int dpaa2_timesync_read_time(struct rte_eth_dev *dev,
@@ -240,5 +237,4 @@ int dpaa2_timesync_read_rx_timestamp(struct rte_eth_dev *dev,
  						uint32_t flags __rte_unused);
  int dpaa2_timesync_read_tx_timestamp(struct rte_eth_dev *dev,
  					  struct timespec *timestamp);
-#endif
  #endif /* _DPAA2_ETHDEV_H */
drivers/net/dpaa2/dpaa2_ptp.c +2 -0
index f58eedb314..899dd5d442 100644
@@ -129,6 +129,7 @@ int dpaa2_timesync_read_rx_timestamp(struct rte_eth_dev *dev,
  	return 0;
  }
 
+#if defined(RTE_LIBRTE_IEEE1588)
  static int
  dpaa2_create_dprtc_device(int vdev_fd __rte_unused,
  			   struct vfio_device_info *obj_info __rte_unused,
@@ -179,3 +180,4 @@ static struct rte_dpaa2_object rte_dpaa2_dprtc_obj = {
  };
 
  RTE_PMD_REGISTER_DPAA2_OBJECT(dprtc, rte_dpaa2_dprtc_obj);
+#endif
drivers/net/dpaa2/meson.build +2 -5
index 6dd0eb274c..4312aa73f7 100644
@@ -13,15 +13,12 @@ sources = files('base/dpaa2_hw_dpni.c',
  		'dpaa2_flow.c',
  		'dpaa2_rxtx.c',
  		'dpaa2_sparser.c',
+		'dpaa2_ptp.c',
+		'mc/dprtc.c',
  		'mc/dpkg.c',
  		'mc/dpdmux.c',
  		'mc/dpni.c')
 
-if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
-	sources += files('mc/dprtc.c')
-	sources += files('dpaa2_ptp.c')
-endif
-
  includes += include_directories('base', 'mc')
 
  install_headers('rte_pmd_dpaa2.h')
#14 lore raw

On 10/8/2020 3:24 AM, Sachin Saxena (OSS) wrote:

[please don't top post, I moved reply to down]

Quoted 108 lines · Ferruh Yigit
Quoted 93 lines

From: Chenxu Di<[email protected]>

Some timesync related source files are built only when 'RTE_LIBRTE_IEEE1588' config is set, which is missing in meson.

Building with '-DRTE_LIBRTE_IEEE1588' cflag will enable timesync functionality in the files that are build, but it won't build files that require 'RTE_LIBRTE_IEEE1588' config.
This causes the build error.

Fixing by removing config check in meson file and build all .c files by default, but wrap relevant code part with 'RTE_LIBRTE_IEEE1588' macro.

Also removing 'RTE_LIBRTE_IEEE1588' ifdef around some fields of the data structures, to not require finer grained macro wraps on the functions using them.
Since the registration of the function disabled with macro check, having functions compiled shouldn't affect the functionality.

Fixes 184c39d16568 ("net/dpaa2: add DPRTC sub-module")
---
  drivers/net/dpaa2/dpaa2_ethdev.h | 4 ----
  drivers/net/dpaa2/dpaa2_ptp.c    | 2 ++
  drivers/net/dpaa2/meson.build    | 7 ++-----
  3 files changed, 4 insertions(+), 9 deletions(-)
drivers/net/dpaa2/dpaa2_ethdev.h +0 -4
index e577355d57..94cf253827 100644
@@ -160,7 +160,6 @@ struct dpaa2_dev_priv {
  	uint16_t ss_offset;
  	uint64_t ss_iova;
  	uint64_t ss_param_iova;
-#if defined(RTE_LIBRTE_IEEE1588)
  	/*stores timestamp of last received packet on dev*/
  	uint64_t rx_timestamp;
  	/*stores timestamp of last received tx confirmation packet on dev*/
@@ -169,7 +168,6 @@ struct dpaa2_dev_priv {
  	 * it corresponds to last packet transmitted
  	 */
  	struct dpaa2_queue *next_tx_conf_queue;
-#endif
 
  	struct rte_eth_dev *eth_dev; /**< Pointer back to holding ethdev */
 
@@ -227,7 +225,6 @@ void dpaa2_dev_free_eqresp_buf(uint16_t eqresp_ci);
  void dpaa2_flow_clean(struct rte_eth_dev *dev);
  uint16_t dpaa2_dev_tx_conf(void *queue)  __rte_unused;
 
-#if defined(RTE_LIBRTE_IEEE1588)
  int dpaa2_timesync_enable(struct rte_eth_dev *dev);
  int dpaa2_timesync_disable(struct rte_eth_dev *dev);
  int dpaa2_timesync_read_time(struct rte_eth_dev *dev,
@@ -240,5 +237,4 @@ int dpaa2_timesync_read_rx_timestamp(struct rte_eth_dev *dev,
  						uint32_t flags __rte_unused);
  int dpaa2_timesync_read_tx_timestamp(struct rte_eth_dev *dev,
  					  struct timespec *timestamp);
-#endif
  #endif /* _DPAA2_ETHDEV_H */
drivers/net/dpaa2/dpaa2_ptp.c +2 -0
index f58eedb314..899dd5d442 100644
@@ -129,6 +129,7 @@ int dpaa2_timesync_read_rx_timestamp(struct rte_eth_dev *dev,
  	return 0;
  }
 
+#if defined(RTE_LIBRTE_IEEE1588)
  static int
  dpaa2_create_dprtc_device(int vdev_fd __rte_unused,
  			   struct vfio_device_info *obj_info __rte_unused,
@@ -179,3 +180,4 @@ static struct rte_dpaa2_object rte_dpaa2_dprtc_obj = {
  };
 
  RTE_PMD_REGISTER_DPAA2_OBJECT(dprtc, rte_dpaa2_dprtc_obj);
+#endif
drivers/net/dpaa2/meson.build +2 -0
index 6dd0eb274c..4312aa73f7 100644
@@ -13,15 +13,12 @@ sources = files('base/dpaa2_hw_dpni.c',
  		'dpaa2_flow.c',
  		'dpaa2_rxtx.c',
  		'dpaa2_sparser.c',
+		'dpaa2_ptp.c',
+		'mc/dprtc.c',
  		'mc/dpkg.c',
  		'mc/dpdmux.c',

drivers/net/dpaa2/meson.build sources = files('base/dpaa2_hw_dpni.c', · line 24

Quoted 10 lines
drivers/net/dpaa2/meson.build +0 -5
  		'mc/dpni.c')
 
-if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
-	sources += files('mc/dprtc.c')
-	sources += files('dpaa2_ptp.c')
-endif
-
  includes += include_directories('base', 'mc')
 
  install_headers('rte_pmd_dpaa2.h')

Following change should also be handled as suggested by Ferruh in earlier.

Quoted 12 lines
    diff --git a/drivers/net/dpaa2/dpaa2_ptp.c b/drivers/net/dpaa2/dpaa2_ptp.c
    index 9f755e84bf..0e44d4c6bf 100644
    --- a/drivers/net/dpaa2/dpaa2_ptp.c
    +++ b/drivers/net/dpaa2/dpaa2_ptp.c
    @@ -178,4 +178,6 @@ static struct rte_dpaa2_object rte_dpaa2_dprtc_obj = {
             .create = dpaa2_create_dprtc_device,
      };

    +#if defined(RTE_LIBRTE_IEEE1588)
      RTE_PMD_REGISTER_DPAA2_OBJECT(dprtc, rte_dpaa2_dprtc_obj);
    +#endif

I think it is already covered in the patch.

Since just wrapping the macro cause a warning on unused static variable/function, I need to wrap 'rte_dpaa2_dprtc_obj' & 'dpaa2_create_dprtc_device()' too along with above macro you mentioned.

#15 lore raw
Quoted 127 lines · Ferruh Yigit

On 10/8/2020 3:24 AM, Sachin Saxena (OSS) wrote:

[please don't top post, I moved reply to down]

Quoted 95 lines · Ferruh Yigit
Quoted 93 lines

From: Chenxu Di<[email protected]>

Some timesync related source files are built only when 'RTE_LIBRTE_IEEE1588' config is set, which is missing in meson.

Building with '-DRTE_LIBRTE_IEEE1588' cflag will enable timesync functionality in the files that are build, but it won't build files that require 'RTE_LIBRTE_IEEE1588' config.
This causes the build error.

Fixing by removing config check in meson file and build all .c files by default, but wrap relevant code part with 'RTE_LIBRTE_IEEE1588' macro.

Also removing 'RTE_LIBRTE_IEEE1588' ifdef around some fields of the data structures, to not require finer grained macro wraps on the functions using them.
Since the registration of the function disabled with macro check, having functions compiled shouldn't affect the functionality.

Fixes 184c39d16568 ("net/dpaa2: add DPRTC sub-module")

---
  drivers/net/dpaa2/dpaa2_ethdev.h | 4 ----
  drivers/net/dpaa2/dpaa2_ptp.c    | 2 ++
  drivers/net/dpaa2/meson.build    | 7 ++-----

  3 files changed, 4 insertions(+), 9 deletions(-)
drivers/net/dpaa2/dpaa2_ethdev.h +0 -2
index e577355d57..94cf253827 100644
@@ -160,7 +160,6 @@ struct dpaa2_dev_priv {
       uint16_t ss_offset;
       uint64_t ss_iova;
       uint64_t ss_param_iova;
-#if defined(RTE_LIBRTE_IEEE1588)
       /*stores timestamp of last received packet on dev*/
       uint64_t rx_timestamp;
       /*stores timestamp of last received tx confirmation packet on dev*/
@@ -169,7 +168,6 @@ struct dpaa2_dev_priv {
        * it corresponds to last packet transmitted
        */
       struct dpaa2_queue *next_tx_conf_queue;
-#endif

        struct rte_eth_dev *eth_dev; /**< Pointer back to holding ethdev */ @@ -227,7 +225,6 @@ void dpaa2_dev_free_eqresp_buf(uint16_t eqresp_ci); void dpaa2_flow_clean(struct rte_eth_dev *dev);
  uint16_t dpaa2_dev_tx_conf(void *queue)  __rte_unused;
  -#if defined(RTE_LIBRTE_IEEE1588)
  int dpaa2_timesync_enable(struct rte_eth_dev *dev);
  int dpaa2_timesync_disable(struct rte_eth_dev *dev);
  int dpaa2_timesync_read_time(struct rte_eth_dev *dev,

drivers/net/dpaa2/dpaa2_ethdev.h +0 -1
@@ -240,5 +237,4 @@ int dpaa2_timesync_read_rx_timestamp(struct 
 rte_eth_dev *dev,
                           uint32_t flags __rte_unused);
   int dpaa2_timesync_read_tx_timestamp(struct rte_eth_dev *dev,
                         struct timespec *timestamp);
-#endif

  #endif /* _DPAA2_ETHDEV_H */

drivers/net/dpaa2/dpaa2_ptp.c +0 -0
index f58eedb314..899dd5d442 100644
@@ -129,6 +129,7 @@ int dpaa2_timesync_read_rx_timestamp(struct 
 rte_eth_dev *dev,
       return 0;
   }
   +#if defined(RTE_LIBRTE_IEEE1588)

  static int
  dpaa2_create_dprtc_device(int vdev_fd __rte_unused, struct vfio_device_info *obj_info __rte_unused,

drivers/net/dpaa2/dpaa2_ptp.c +1 -0
@@ -179,3 +180,4 @@ static struct rte_dpaa2_object 
 rte_dpaa2_dprtc_obj = {
   };
     RTE_PMD_REGISTER_DPAA2_OBJECT(dprtc, rte_dpaa2_dprtc_obj);
+#endif
drivers/net/dpaa2/meson.build +2 -0
index 6dd0eb274c..4312aa73f7 100644
@@ -13,15 +13,12 @@ sources = files('base/dpaa2_hw_dpni.c',
           'dpaa2_flow.c',
           'dpaa2_rxtx.c',
           'dpaa2_sparser.c',
+        'dpaa2_ptp.c',
+        'mc/dprtc.c',
           'mc/dpkg.c',
           'mc/dpdmux.c',
           'mc/dpni.c')
   -if dpdk_conf.has('RTE_LIBRTE_IEEE1588')
Quoted 8 lines

drivers/net/dpaa2/meson.build sources = files('base/dpaa2_hw_dpni.c', · line 24

Quoted 6 lines
drivers/net/dpaa2/meson.build +0 -4
-    sources += files('mc/dprtc.c')
-    sources += files('dpaa2_ptp.c')
-endif
-
   includes += include_directories('base', 'mc')
     install_headers('rte_pmd_dpaa2.h')

Following change should also be handled as suggested by Ferruh in

earlier.

Quoted 2 lines

     diff --git a/drivers/net/dpaa2/dpaa2_ptp.c

b/drivers/net/dpaa2/dpaa2_ptp.c

Quoted 4 lines

     index 9f755e84bf..0e44d4c6bf 100644
     --- a/drivers/net/dpaa2/dpaa2_ptp.c
     +++ b/drivers/net/dpaa2/dpaa2_ptp.c
     @@ -178,4 +178,6 @@ static struct rte_dpaa2_object

rte_dpaa2_dprtc_obj = {

Quoted 7 lines

              .create = dpaa2_create_dprtc_device, };

     +#if defined(RTE_LIBRTE_IEEE1588) RTE_PMD_REGISTER_DPAA2_OBJECT(dprtc, rte_dpaa2_dprtc_obj); +#endif

I think it is already covered in the patch.

Since just wrapping the macro cause a warning on unused static variable/function, I need to wrap 'rte_dpaa2_dprtc_obj' & 'dpaa2_create_dprtc_device()' too along with above macro you mentioned.

Applied
#16 lore raw
Quoted 26 lines · Sachin Saxena (OSS)
Quoted 25 lines · Ferruh Yigit

On 10/8/2020 3:24 AM, Sachin Saxena (OSS) wrote:

[please don't top post, I moved reply to down]

Quoted 21 lines · Ferruh Yigit
Quoted 19 lines

From: Chenxu Di<[email protected]>

Some timesync related source files are built only when 'RTE_LIBRTE_IEEE1588' config is set, which is missing in meson.

Building with '-DRTE_LIBRTE_IEEE1588' cflag will enable timesync functionality in the files that are build, but it won't build files that require 'RTE_LIBRTE_IEEE1588' config.
This causes the build error.

Fixing by removing config check in meson file and build all .c files by default, but wrap relevant code part with 'RTE_LIBRTE_IEEE1588' macro.

Also removing 'RTE_LIBRTE_IEEE1588' ifdef around some fields of the data structures, to not require finer grained macro wraps on the functions using them.
Since the registration of the function disabled with macro check, having functions compiled shouldn't affect the functionality.

Quoted 8 lines
Quoted 5 lines
Quoted 5 lines
Quoted 5 lines
Fixes 184c39d16568 ("net/dpaa2: add DPRTC sub-module")

Applied to dpdk-next-net/main, thanks.