Skip to contentSkip to conversation

build fixes on FreeBSD

PATCH Applied Bruce Richardson 3 participants 9 messages Mar 11, 2022 – Mar 15, 2022 Arm/Amlogic Meson SoC support

lore
Share & tools
lore
#0 lore raw

cover build fixes on FreeBSD

This set contains a number of fixes, most of which are for build issues discovered when compiling on FreeBSD with clang.

Bruce Richardson (5):
eal/freebsd: add missing C++ include guards
compressdev: separate out driver-only headers
compressdev: fix missing space in header
cryptodev: fix compilation with clang C++ builds
eventdev: fix compilation with clang C++ builds

 lib/compressdev/meson.build                |  6 ++++--
 lib/compressdev/rte_compressdev_internal.h |  2 +-
 lib/cryptodev/rte_crypto.h                 |  9 +++++++++
 lib/eal/freebsd/include/rte_os.h           |  8 ++++++++
 lib/eventdev/rte_eventdev.h                | 12 +++++++++++-
 5 files changed, 33 insertions(+), 4 deletions(-)

--
2.34.1

#1 lore raw

PATCH 1/5 eal/freebsd: add missing C++ include guards

Add mising 'extern "C"' to file.

Fixes 428eb983f5f7 ("eal: add OS specific header file")
---
 lib/eal/freebsd/include/rte_os.h | 8 ++++++++
 1 file changed, 8 insertions(+)
lib/eal/freebsd/include/rte_os.h +8 -0
index 9d8a69008c..b4afd45adc 100644
@@ -5,6 +5,10 @@
 #ifndef _RTE_OS_H_
 #define _RTE_OS_H_
 
+#ifdef __cplusplus
+extern "C" {
+#endif
+
 /**
  * This header should contain any definition
  * which is not supported natively or named differently in FreeBSD.
@@ -59,4 +63,8 @@ typedef cpuset_t rte_cpuset_t;
 } while (0)
 #endif
 
+#ifdef __cplusplus
+}
+#endif
+
 #endif /* _RTE_OS_H_ */
Signature · 1 line
2.34.1
#2 lore raw

PATCH 2/5 compressdev: separate out driver-only headers

In reply to Bruce Richardson

The headers rte_compressdev_pmd.h and rte_compressdev_internal.h are, as the filenames suggest, headers for building drivers using the compressdev APIs. As such they should be marked as "driver_sdk_headers" rather than just "headers" in the meson.build file.

---
 lib/compressdev/meson.build | 6 ++++--
 1 file changed, 4 insertions(+), 2 deletions(-)
lib/compressdev/meson.build +4 -2
index 75ea666a9b..c80295dc0d 100644
@@ -11,7 +11,9 @@ sources = files('rte_compressdev.c',
     'rte_compressdev_pmd.c',
     'rte_comp.c')
 headers = files('rte_compressdev.h',
-    'rte_compressdev_pmd.h',
-    'rte_compressdev_internal.h',
     'rte_comp.h')
+driver_sdk_headers = files(
+        'rte_compressdev_pmd.h',
+        'rte_compressdev_internal.h',
+    )
 deps += ['kvargs', 'mbuf']
Signature · 1 line
2.34.1
#3 lore raw

PATCH 3/5 compressdev: fix missing space in header

In reply to Bruce Richardson

Building with clang on FreeBSD with chkincs enabled, we get the following error about a missing space:

../lib/compressdev/rte_compressdev_internal.h:25:58: error: invalid suffix on literal; C++11 requires a space between literal and identifier [-Wreserved-user-defined-literal]

        rte_log(RTE_LOG_ ## level, compressdev_logtype, "%s(): "fmt "\n", \

Adding in a space between the '"' and 'fmt' removes the error.

Fixes ed7dd94f7f66 ("compressdev: add basic device management")
---
 lib/compressdev/rte_compressdev_internal.h | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)
lib/compressdev/rte_compressdev_internal.h +1 -1
index 25d8afbfb9..b3b193e3ee 100644
@@ -22,7 +22,7 @@ extern "C" {
 /* Logging Macros */
 extern int compressdev_logtype;
 #define COMPRESSDEV_LOG(level, fmt, args...) \
-	rte_log(RTE_LOG_ ## level, compressdev_logtype, "%s(): "fmt "\n", \
+	rte_log(RTE_LOG_ ## level, compressdev_logtype, "%s(): " fmt "\n", \
 			__func__, ##args)
 
 /**
Signature · 1 line
2.34.1
#4 lore raw

PATCH 4/5 cryptodev: fix compilation with clang C++ builds

In reply to Bruce Richardson

When compiling on FreeBSD with clang and include checking enabled, errors are emitted due to differences in how empty structs/unions are handled in C and C++, as C++ structs cannot have zero size.

../lib/cryptodev/rte_crypto.h:127:2: error: union has size 0 in C, non-zero size in C++

Since the contents of the union are all themselves of zero size, the actual union wrapper is unnecessary. We therefore remove it for C++ builds - though keep it for C builds for safety and clarity of understanding the code.

Fixes c0f87eb5252b ("cryptodev: change burst API to be crypto op oriented")
Fixes d2a4223c4c6d ("cryptodev: do not store pointer to op specific params")
---
 lib/cryptodev/rte_crypto.h | 9 +++++++++
 1 file changed, 9 insertions(+)
lib/cryptodev/rte_crypto.h +9 -0
index a864f5036f..aeb3bf6e38 100644
@@ -123,15 +123,24 @@ struct rte_crypto_op {
 	rte_iova_t phys_addr;
 	/**< physical address of crypto operation */
 
+/* empty structures do not have zero size in C++ leading to compilation errors
+ * with clang about structure/union having different sizes in C and C++.
+ * While things are clearer with an explicit union, since each field is
+ * zero-sized it's not actually needed, so omit it for C++
+ */
+#ifndef __cplusplus
 	__extension__
 	union {
+#endif
 		struct rte_crypto_sym_op sym[0];
 		/**< Symmetric operation parameters */
 
 		struct rte_crypto_asym_op asym[0];
 		/**< Asymmetric operation parameters */
 
+#ifndef __cplusplus
 	}; /**< operation specific parameters */
+#endif
 };
 
 /**
Signature · 1 line
2.34.1
#5 lore raw

PATCH 5/5 eventdev: fix compilation with clang C++ builds

In reply to Bruce Richardson

When compiling on FreeBSD with clang and include checking enabled, errors are emitted due to differences in how empty structs/unions are handled in C and C++, as C++ structs cannot have zero size.

../lib/eventdev/rte_eventdev.h:992:2: error: union has size 0 in C, non-zero size in C++

Since the contents of the union are all themselves of zero size, the actual union wrapper is unnecessary. We therefore remove it for C++ builds - though keep it for C builds for safety and clarity of understanding the code. The alignment constraint on the union is unnecessary in the case where the whole struct is aligned on a 16-byte boundary, so we add that constraint to the overall structure to ensure it applies for C++ code as well as C.

Fixes 1cc44d409271 ("eventdev: introduce event vector capability")
---
 lib/eventdev/rte_eventdev.h | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)
lib/eventdev/rte_eventdev.h +11 -1
index 67c4a5e036..42a5660169 100644
@@ -984,21 +984,31 @@ struct rte_event_vector {
 	};
 	/**< Union to hold common attributes of the vector array. */
 	uint64_t impl_opaque;
+
+/* empty structures do not have zero size in C++ leading to compilation errors
+ * with clang about structure having different sizes in C and C++.
+ * Since these are all zero-sized arrays, we can omit the "union" wrapper for
+ * C++ builds, removing the warning.
+ */
+#ifndef __cplusplus
 	/**< Implementation specific opaque value.
 	 * An implementation may use this field to hold implementation specific
 	 * value to share between dequeue and enqueue operation.
 	 * The application should not modify this field.
 	 */
 	union {
+#endif
 		struct rte_mbuf *mbufs[0];
 		void *ptrs[0];
 		uint64_t *u64s[0];
+#ifndef __cplusplus
 	} __rte_aligned(16);
+#endif
 	/**< Start of the vector array union. Depending upon the event type the
 	 * vector array can be an array of mbufs or pointers or opaque u64
 	 * values.
 	 */
-};
+} __rte_aligned(16);
 
 /* Scheduler type definitions */
 #define RTE_SCHED_TYPE_ORDERED          0
Signature · 1 line
2.34.1
#6 lore raw

Re: [PATCH 5/5] eventdev: fix compilation with clang C++ builds

Quoted 49 lines · On Fri, 11 Mar 2022 20:05:23 +0000 Bruce Richardson…

When compiling on FreeBSD with clang and include checking enabled, errors are emitted due to differences in how empty structs/unions are handled in C and C++, as C++ structs cannot have zero size.

../lib/eventdev/rte_eventdev.h:992:2: error: union has size 0 in C, non-zero size in C++

Since the contents of the union are all themselves of zero size, the actual union wrapper is unnecessary. We therefore remove it for C++ builds - though keep it for C builds for safety and clarity of understanding the code. The alignment constraint on the union is unnecessary in the case where the whole struct is aligned on a 16-byte boundary, so we add that constraint to the overall structure to ensure it applies for C++ code as well as C.

Fixes 1cc44d409271 ("eventdev: introduce event vector capability")
---
 lib/eventdev/rte_eventdev.h | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)
lib/eventdev/rte_eventdev.h +9 -0
index 67c4a5e036..42a5660169 100644
@@ -984,21 +984,31 @@ struct rte_event_vector {
 	};
 	/**< Union to hold common attributes of the vector array. */
 	uint64_t impl_opaque;
+
+/* empty structures do not have zero size in C++ leading to compilation errors
+ * with clang about structure having different sizes in C and C++.
+ * Since these are all zero-sized arrays, we can omit the "union" wrapper for
+ * C++ builds, removing the warning.
+ */
+#ifndef __cplusplus
 	/**< Implementation specific opaque value.
 	 * An implementation may use this field to hold implementation specific
 	 * value to share between dequeue and enqueue operation.
 	 * The application should not modify this field.
 	 */
 	union {
+#endif
 		struct rte_mbuf *mbufs[0];
 		void *ptrs[0];
 		uint64_t *u64s[0];
+#ifndef __cplusplus
 	} __rte_aligned(16);

lib/eventdev/rte_eventdev.h struct rte_event_vector { · line 1014

Quoted 10 lines
lib/eventdev/rte_eventdev.h +2 -1
+#endif
 	/**< Start of the vector array union. Depending upon the event type the
 	 * vector array can be an array of mbufs or pointers or opaque u64
 	 * values.
 	 */
-};
+} __rte_aligned(16);
 
 /* Scheduler type definitions */
 #define RTE_SCHED_TYPE_ORDERED          0

Zero size arrays should be replaced by flexible arrays. Linux has this coccinelle script for that:

// SPDX-License-Identifier: GPL-2.0-only
///
/// Zero-length and one-element arrays are deprecated, see /// Documentation/process/deprecated.rst
/// Flexible-array members should be used instead. ///
//
// Confidence: High
// Copyright: (C) 2020 Denis Efremov ISPRAS.
// Comments:
// Options: --no-includes --include-headers

virtual context
virtual report
virtual org
virtual patch

@initialize:python@
@@
def relevant(positions):
for p in positions:
if "uapi" in p.file:
return False
return True

@r depends on !patch@
identifier name, array;
type T;
position p : script:python() { relevant(p) };
@@

(

  struct name {
    ...
*   T array@p[\(0\|1\)];
  };
|
  struct {
    ...
*   T array@p[\(0\|1\)];
  };
|
  union name {
    ...
*   T array@p[\(0\|1\)];
  };
|
  union {
    ...

* T array@p[\(0\|1\)];
};
)

@only_field depends on patch@
identifier name, array;
type T;
position q;
@@

(
struct name {@q

    T array[0];
  };
|
  struct {@q
    T array[0];

};
)

@depends on patch@
identifier name, array;
type T;
position p : script:python() { relevant(p) };
// position @q with rule "only_field" simplifies
// handling of bitfields, arrays, etc.
position q != only_field.q;
@@

(
struct name {@q

    ...
    T array@p[
-       0
    ];
  };
|
  struct {@q
    ...
    T array@p[
-       0
    ];

};
)

@script: python depends on report@
p << r.p;
@@

msg = "WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)"
coccilib.report.print_report(p[0], msg)

@script: python depends on org@
p << r.p;
@@

msg = "WARNING use flexible-array member instead (https://www.kernel.org/doc/html/latest/process/deprecated.html#zero-length-and-one-element-arrays)"
coccilib.org.print_todo(p[0], msg)

#7 lore raw

Re: [PATCH 5/5] eventdev: fix compilation with clang C++ builds

Quoted 60 lines · Stephen Hemminger
Quoted 57 lines · On Fri, 11 Mar 2022 20:05:23 +0000 Bruce Richardson…

When compiling on FreeBSD with clang and include checking enabled, errors are emitted due to differences in how empty structs/unions are handled in C and C++, as C++ structs cannot have zero size.

../lib/eventdev/rte_eventdev.h:992:2: error: union has size 0 in C, non-zero size in C++

Since the contents of the union are all themselves of zero size, the actual union wrapper is unnecessary. We therefore remove it for C++ builds - though keep it for C builds for safety and clarity of understanding the code. The alignment constraint on the union is unnecessary in the case where the whole struct is aligned on a 16-byte boundary, so we add that constraint to the overall structure to ensure it applies for C++ code as well as C.

Fixes 1cc44d409271 ("eventdev: introduce event vector capability")
---
 lib/eventdev/rte_eventdev.h | 12 +++++++++++-
 1 file changed, 11 insertions(+), 1 deletion(-)
lib/eventdev/rte_eventdev.h +11 -1
index 67c4a5e036..42a5660169 100644
@@ -984,21 +984,31 @@ struct rte_event_vector {
 	};
 	/**< Union to hold common attributes of the vector array. */
 	uint64_t impl_opaque;
+
+/* empty structures do not have zero size in C++ leading to compilation errors
+ * with clang about structure having different sizes in C and C++.
+ * Since these are all zero-sized arrays, we can omit the "union" wrapper for
+ * C++ builds, removing the warning.
+ */
+#ifndef __cplusplus
 	/**< Implementation specific opaque value.
 	 * An implementation may use this field to hold implementation specific
 	 * value to share between dequeue and enqueue operation.
 	 * The application should not modify this field.
 	 */
 	union {
+#endif
 		struct rte_mbuf *mbufs[0];
 		void *ptrs[0];
 		uint64_t *u64s[0];
+#ifndef __cplusplus
 	} __rte_aligned(16);
+#endif
 	/**< Start of the vector array union. Depending upon the event type the
 	 * vector array can be an array of mbufs or pointers or opaque u64
 	 * values.
 	 */
-};
+} __rte_aligned(16);
 
Quoted 6 lines

lib/eventdev/rte_eventdev.h struct rte_event_vector { · line 1014

Quoted 2 lines
lib/eventdev/rte_eventdev.h +0 -0
 /* Scheduler type definitions */
 #define RTE_SCHED_TYPE_ORDERED          0

Zero size arrays should be replaced by flexible arrays. Linux has this coccinelle script for that:

Yes, I agree. However, given how late in the release process I discovered this, I wanted a fix with absolute minimum impact, which is why I chose the above. Zero struct change for C code, and enough of a fix for C++ to get it compiling.
If you feel that replacing with flexible arrays is safe enough to do at this late stage, we can perhaps consider it, but overall I'd suggest doing any such replacement in 22.07

/Bruce

#8 lore raw
In reply to Bruce Richardson

11/03/2022 21:05, Bruce Richardson:

Quoted 9 lines

This set contains a number of fixes, most of which are for build issues discovered when compiling on FreeBSD with clang.

Bruce Richardson (5):
eal/freebsd: add missing C++ include guards
compressdev: separate out driver-only headers
compressdev: fix missing space in header
cryptodev: fix compilation with clang C++ builds
eventdev: fix compilation with clang C++ builds

Applied, thanks.