test/compress: fix coverity warnings
PATCH 1/1 test/compress: fix coverity warnings
CID 340857: Null pointer dereferences (NULL_RETURNS)
CID 340856: (CONSTANT_EXPRESSION_RESULT)
--- app/test/test_compressdev.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-)
app/test/test_compressdev.c +8 -2
@@ -755,6 +755,12 @@ test_deflate_comp_decomp(const struct interim_data_params *int_data,char *contig_buf =NULL ;uint64_t compress_checksum[num_bufs]; +if (capa ==NULL ) { +RTE_LOG (ERR ,USER1 , +"Compress device does not support DEFLATE\n" ); +return -1 ; + } +/* Initialize all arrays to NULL */ memset (uncomp_bufs,0 ,sizeof (struct rte_mbuf *) * num_bufs);memset (comp_bufs,0 ,sizeof (struct rte_mbuf *) * num_bufs);
@@ -975,7 +981,7 @@ test_deflate_comp_decomp(const struct interim_data_params *int_data,enum rte_comp_huffman huffman_type = compress_xform->deflate.huffman;char engine[] ="zlib (directly, not PMD)" ; -if (zlib_dir !=ZLIB_COMPRESS || zlib_dir !=ZLIB_ALL ) +if (zlib_dir !=ZLIB_COMPRESS && zlib_dir !=ZLIB_ALL )strlcpy (engine,"PMD" ,sizeof (engine)); RTE_LOG(DEBUG, USER1, "Buffer %u compressed by %s from %u to"
@@ -1211,7 +1217,7 @@ test_deflate_comp_decomp(const struct interim_data_params *int_data,for (i =0 ; i < num_bufs; i++) { priv_data = (struct priv_op_data *)(ops_processed[i] +1 );char engine[] ="zlib, (directly, no PMD)" ; -if (zlib_dir !=ZLIB_DECOMPRESS || zlib_dir !=ZLIB_ALL ) +if (zlib_dir !=ZLIB_DECOMPRESS && zlib_dir !=ZLIB_ALL )strlcpy (engine,"pmd" ,sizeof (engine)); RTE_LOG(DEBUG, USER1, "Buffer %u decompressed by %s from %u to %u bytes\n",
Signature · 1 line
2.17.1
Hi Adam,
Quoted 23 lines
-----Original Message-----
From: Dybkowski, AdamX
Sent: Tuesday, July 23, 2019 11:54 AM
To: [email protected]; Trahe, Fiona <[email protected]>; [email protected]
Cc: Dybkowski, AdamX <[email protected]>
Subject: [PATCH] test/compress: fix coverity warningsCID 340857: Null pointer dereferences (
NULL_RETURNS)CID 340856: (
CONSTANT_EXPRESSION_RESULT)Fixes 3be12ea52ad8 ("test/compress: improve debug trace setup")Signed-off-by Adam Dybkowski <[email protected]>--- app/test/test_compressdev.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-)app/test/test_compressdev.c +0
-0
app/test/test_compressdev.c test_deflate_comp_decomp(const struct interim_data_params *int_data, · line 761
Quoted 7 lines
app/test/test_compressdev.c +4
-0@@ -755,6 +755,12 @@ test_deflate_comp_decomp(const struct interim_data_params *int_data,char *contig_buf =NULL ;uint64_t compress_checksum[num_bufs]; +if (capa ==NULL ) { +RTE_LOG (ERR ,USER1 , +"Compress device does not support DEFLATE\n" ); +return -1 ;
[Fiona] shouldn't you return -ENOTSUP?
Hi Fiona.
Quoted 35 lines
-----Original Message-----
From: Trahe, Fiona
Sent: Thursday, 25 July, 2019 15:56
To: Dybkowski, AdamX <[email protected]>; [email protected]; [email protected]
Cc: Trahe, Fiona <[email protected]>
Subject: RE: [PATCH] test/compress: fix coverity warningsHi Adam,
app/test/test_compressdev.c test_deflate_comp_decomp(const struct
Quoted 24 lines
-----Original Message-----
From: Dybkowski, AdamX
Sent: Tuesday, July 23, 2019 11:54 AM
To: [email protected]; Trahe, Fiona <[email protected]>; [email protected]
Cc: Dybkowski, AdamX <[email protected]>
Subject: [PATCH] test/compress: fix coverity warningsCID 340857: Null pointer dereferences (
NULL_RETURNS)CID 340856: (
CONSTANT_EXPRESSION_RESULT)Fixes 3be12ea52ad8 ("test/compress: improve debug trace setup")Signed-off-by Adam Dybkowski <[email protected]>--- app/test/test_compressdev.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-)app/test/test_compressdev.c +0
-0@@ -755,6 +755,12 @@ test_deflate_comp_decomp(const struct
interim_data_params*int_data,
Quoted 8 lines
app/test/test_compressdev.c test_deflate_comp_decomp(const struct interim_data_params *int_data, · line 761
Quoted 7 lines
app/test/test_compressdev.c +4
-0char *contig_buf =NULL ;uint64_t compress_checksum[num_bufs]; +if (capa ==NULL ) { +RTE_LOG (ERR ,USER1 , +"Compress device does not support DEFLATE\n" ); +return -1 ;[Fiona] shouldn't you return
-ENOTSUP?
[Adam] All other errors end with returning -1 (see the ret_status variable which is initialized to -1). That's why I handle the error case the same way. This is inside the test suite, not a PMD.
Trahe, Fiona
Acked-byHi Adam,
Quoted 46 lines
-----Original Message-----
From: Dybkowski, AdamX
Sent: Thursday, July 25, 2019 3:23 PM
To: Trahe, Fiona <[email protected]>; [email protected]; [email protected]
Subject: RE: [PATCH] test/compress: fix coverity warningsHi Fiona.
Quoted 38 lines
-----Original Message-----
From: Trahe, Fiona
Sent: Thursday, 25 July, 2019 15:56
To: Dybkowski, AdamX <[email protected]>; [email protected]; [email protected]
Cc: Trahe, Fiona <[email protected]>
Subject: RE: [PATCH] test/compress: fix coverity warningsHi Adam,
app/test/test_compressdev.c test_deflate_comp_decomp(const struct
Quoted 24 lines
-----Original Message-----
From: Dybkowski, AdamX
Sent: Tuesday, July 23, 2019 11:54 AM
To: [email protected]; Trahe, Fiona <[email protected]>; [email protected]
Cc: Dybkowski, AdamX <[email protected]>
Subject: [PATCH] test/compress: fix coverity warningsCID 340857: Null pointer dereferences (
NULL_RETURNS)CID 340856: (
CONSTANT_EXPRESSION_RESULT)Fixes 3be12ea52ad8 ("test/compress: improve debug trace setup")Signed-off-by Adam Dybkowski <[email protected]>--- app/test/test_compressdev.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-)app/test/test_compressdev.c +0
-0@@ -755,6 +755,12 @@ test_deflate_comp_decomp(const struct
interim_data_params*int_data,Quoted 3 lines
app/test/test_compressdev.c +0
-0char *contig_buf =NULL ;uint64_t compress_checksum[num_bufs];
Quoted 7 lines
Quoted 5 lines
app/test/test_compressdev.c test_deflate_comp_decomp(const struct interim_data_params *int_data, · line 761
Quoted 4 lines
app/test/test_compressdev.c +4
-0+if (capa ==NULL ) { +RTE_LOG (ERR ,USER1 , +"Compress device does not support DEFLATE\n" ); +return -1 ;[Fiona] shouldn't you return
-ENOTSUP?[Adam] All other errors end with returning -1 (see the
ret_statusvariable which is initialized to -1). That's why I handle the error case the same way. This is inside the test suite, not a PMD.
[Fiona] ok, can see in the code that the capability is already checked at the higher level for this case, so no need to return ENOTSUP here. -1 is ok
Akhil Goyal
AppliedQuoted 26 lines
Quoted 26 lines
Quoted 26 lines
app/test/test_compressdev.c test_deflate_comp_decomp(const struct
Quoted 17 lines
CID 340857: Null pointer dereferences (
NULL_RETURNS)CID 340856: (
CONSTANT_EXPRESSION_RESULT)Fixes 3be12ea52ad8 ("test/compress: improve debug trace setup")Signed-off-by Adam Dybkowski <[email protected]>--- app/test/test_compressdev.c | 10 ++++++++-- 1 file changed, 8 insertions(+), 2 deletions(-)app/test/test_compressdev.c +0
-0@@ -755,6 +755,12 @@ test_deflate_comp_decomp(const struct
interim_data_params*int_data,app/test/test_compressdev.c test_deflate_comp_decomp(const struct interim_data_params *int_data, · line 761
Quoted 7 lines
app/test/test_compressdev.c +4
-0char *contig_buf =NULL ;uint64_t compress_checksum[num_bufs]; +if (capa ==NULL ) { +RTE_LOG (ERR ,USER1 , +"Compress device does not support DEFLATE\n" ); +return -1 ;[Fiona] shouldn't you return
-ENOTSUP?
Quoted 8 lines
Quoted 1 line
[Adam] All other errors end with returning -1 (see the
ret_statusvariable whichis initialized to -1).
Quoted 1 line
That's why I handle the error case the same way. This is inside the test suite,
not a PMD.
[Fiona] ok, can see in the code that the capability is already checked at the higher level
for this case, so no need to return ENOTSUP here. -1 is okAcked-by Fiona Trahe <[email protected]>
Applied to dpdk-next-crypto
Thanks.