* [PATCH dwarves 0/2] flexible_arrays tests fixes @ 2024-10-24 11:26 Alan Maguire 2024-10-24 11:26 ` [PATCH dwarves 1/2] tests/flexible_arrays: redirect stderr to avoid warnings Alan Maguire 2024-10-24 11:26 ` [PATCH dwarves 2/2] tests/flexible_arrays: print flex array struct from same vmlinux Alan Maguire 0 siblings, 2 replies; 5+ messages in thread From: Alan Maguire @ 2024-10-24 11:26 UTC (permalink / raw) To: acme; +Cc: dwarves, Alan Maguire Two small fixes for the flex arrays tests; redirect stderr to avoid seeing warnings during tests (patch 1), and ensure we are printing flex arrays from the same vmlinux (patch 2). Alan Maguire (2): tests/flexible_arrays: redirect stderr to avoid warnings tests/flexible_arrays: print flex array struct from same vmlinux tests/flexible_arrays.sh | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) -- 2.43.5 ^ permalink raw reply [flat|nested] 5+ messages in thread
* [PATCH dwarves 1/2] tests/flexible_arrays: redirect stderr to avoid warnings 2024-10-24 11:26 [PATCH dwarves 0/2] flexible_arrays tests fixes Alan Maguire @ 2024-10-24 11:26 ` Alan Maguire [not found] ` <CA+JHD923mvOrr9JoKeVkeB1k2PTGSHsi7=6HPrV4tVNF1R=GWA@mail.gmail.com> 2024-10-24 11:26 ` [PATCH dwarves 2/2] tests/flexible_arrays: print flex array struct from same vmlinux Alan Maguire 1 sibling, 1 reply; 5+ messages in thread From: Alan Maguire @ 2024-10-24 11:26 UTC (permalink / raw) To: acme; +Cc: dwarves, Alan Maguire With recent kernel, saw 2: Flexible arrays accounting: WARNING: still unsuported BTF_KIND_DECL_TAG(bpf_fastcall) for bpf_cast_to_kern_ctx already with attribute (bpf_kfunc), ignoring WARNING: still unsuported BTF_KIND_DECL_TAG(bpf_fastcall) for bpf_rdonly_cast already with attribute (bpf_kfunc), ignoring Ok Redirect pahole output to avoid seeing the stderr messaging. Signed-off-by: Alan Maguire <alan.maguire@oracle.com> --- tests/flexible_arrays.sh | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) diff --git a/tests/flexible_arrays.sh b/tests/flexible_arrays.sh index 7c21253..e9e3cb0 100755 --- a/tests/flexible_arrays.sh +++ b/tests/flexible_arrays.sh @@ -20,7 +20,7 @@ pretty=$(mktemp /tmp/flexible_arrays.data.sh.XXXXXX.c) echo -n "Flexible arrays accounting: " -for struct in $(pahole -F btf --sizes --with_embedded_flexible_array $vmlinux | cut -f1) ; do +for struct in $(pahole -F btf --sizes --with_embedded_flexible_array $vmlinux 2>/dev/null| cut -f1) ; do pahole $struct > $pretty # We need to check for just one tab before the comment as when expanding unnamed -- 2.43.5 ^ permalink raw reply related [flat|nested] 5+ messages in thread
[parent not found: <CA+JHD923mvOrr9JoKeVkeB1k2PTGSHsi7=6HPrV4tVNF1R=GWA@mail.gmail.com>]
* Re: [PATCH dwarves 1/2] tests/flexible_arrays: redirect stderr to avoid warnings [not found] ` <CA+JHD923mvOrr9JoKeVkeB1k2PTGSHsi7=6HPrV4tVNF1R=GWA@mail.gmail.com> @ 2024-10-24 13:01 ` Alan Maguire 2024-10-24 13:45 ` Arnaldo Carvalho de Melo 0 siblings, 1 reply; 5+ messages in thread From: Alan Maguire @ 2024-10-24 13:01 UTC (permalink / raw) To: Arnaldo Carvalho de Melo; +Cc: Arnaldo Carvalho de Melo, dwarves On 24/10/2024 13:13, Arnaldo Carvalho de Melo wrote: > > On Thu, Oct 24, 2024, 8:26 AM Alan Maguire <alan.maguire@oracle.com > <mailto:alan.maguire@oracle.com>> wrote: > > With recent kernel, saw > > 2: Flexible arrays accounting: WARNING: still unsuported > BTF_KIND_DECL_TAG(bpf_fastcall) for bpf_cast_to_kern_ctx already > with attribute (bpf_kfunc), ignoring > WARNING: still unsuported BTF_KIND_DECL_TAG(bpf_fastcall) for > bpf_rdonly_cast already with attribute (bpf_kfunc), ignoring > Ok > > Redirect pahole output to avoid seeing the stderr messaging. > > > > This is something I kinda expect from these tests, now I want to get > hold of such a vmlinux to understand what's needed to address the warning. > > Maybe have a quiet mode for testing just what's supported in a released > version when tested by packagers? > > Maybe not, as knowing that they have a kernel with a feature unsupported > by pahole may be important and should elicit measures to either get a > newer version of pahole or disable a still experimental feature in their > kernel? > > WDYT? > Good point on the visibility of warnings being useful. We have VERBOSE already so it seems like having QUIET (implying stderr > /dev/null) would be easiest to add at this stage rather than having to add multiple levels to VERBOSE. Maybe QUIET isn't even needed though as the above is still clearly a test pass, and the messaging doesn't interfere with test success. Probably best to drop this one for now anyway.. > > > Signed-off -by: Alan Maguire <alan.maguire@oracle.com > <mailto:alan.maguire@oracle.com>> > --- > tests/flexible_arrays.sh | 2 +- > 1 file changed, 1 insertion(+), 1 deletion(-) > > diff --git a/tests/flexible_arrays.sh b/tests/flexible_arrays.sh > index 7c21253..e9e3cb0 100755 > --- a/tests/flexible_arrays.sh > +++ b/tests/flexible_arrays.sh > @@ -20,7 +20,7 @@ pretty=$(mktemp /tmp/flexible_arrays.data.sh > <http://flexible_arrays.data.sh>.XXXXXX.c) > > echo -n "Flexible arrays accounting: " > > -for struct in $(pahole -F btf --sizes -- > with_embedded_flexible_array $vmlinux | cut -f1) ; do > +for struct in $(pahole -F btf --sizes -- > with_embedded_flexible_array $vmlinux 2>/dev/null| cut -f1) ; do > pahole $struct > $pretty > > # We need to check for just one tab before the comment as > when expanding unnamed > -- > 2.43.5 > ^ permalink raw reply [flat|nested] 5+ messages in thread
* Re: [PATCH dwarves 1/2] tests/flexible_arrays: redirect stderr to avoid warnings 2024-10-24 13:01 ` Alan Maguire @ 2024-10-24 13:45 ` Arnaldo Carvalho de Melo 0 siblings, 0 replies; 5+ messages in thread From: Arnaldo Carvalho de Melo @ 2024-10-24 13:45 UTC (permalink / raw) To: Alan Maguire; +Cc: Arnaldo Carvalho de Melo, dwarves On Thu, Oct 24, 2024 at 02:01:21PM +0100, Alan Maguire wrote: > On 24/10/2024 13:13, Arnaldo Carvalho de Melo wrote: > > > > On Thu, Oct 24, 2024, 8:26 AM Alan Maguire <alan.maguire@oracle.com > > <mailto:alan.maguire@oracle.com>> wrote: > > > > With recent kernel, saw > > > > 2: Flexible arrays accounting: WARNING: still unsuported > > BTF_KIND_DECL_TAG(bpf_fastcall) for bpf_cast_to_kern_ctx already > > with attribute (bpf_kfunc), ignoring > > WARNING: still unsuported BTF_KIND_DECL_TAG(bpf_fastcall) for > > bpf_rdonly_cast already with attribute (bpf_kfunc), ignoring > > Ok > > > > Redirect pahole output to avoid seeing the stderr messaging. > > > > > > > > This is something I kinda expect from these tests, now I want to get > > hold of such a vmlinux to understand what's needed to address the warning. > > > > Maybe have a quiet mode for testing just what's supported in a released > > version when tested by packagers? > > > > Maybe not, as knowing that they have a kernel with a feature unsupported > > by pahole may be important and should elicit measures to either get a > > newer version of pahole or disable a still experimental feature in their > > kernel? > > > > WDYT? > > > > Good point on the visibility of warnings being useful. We have VERBOSE > already so it seems like having QUIET (implying stderr > /dev/null) > would be easiest to add at this stage rather than having to add multiple > levels to VERBOSE. Maybe QUIET isn't even needed though as the above is > still clearly a test pass, and the messaging doesn't interfere with test > success. > > Probably best to drop this one for now anyway.. That is what I did, agreed. I adapted the second one to this fact and applied it, thanks! - Arnaldo > > > > > > Signed-off -by: Alan Maguire <alan.maguire@oracle.com > > <mailto:alan.maguire@oracle.com>> > > --- > > tests/flexible_arrays.sh | 2 +- > > 1 file changed, 1 insertion(+), 1 deletion(-) > > > > diff --git a/tests/flexible_arrays.sh b/tests/flexible_arrays.sh > > index 7c21253..e9e3cb0 100755 > > --- a/tests/flexible_arrays.sh > > +++ b/tests/flexible_arrays.sh > > @@ -20,7 +20,7 @@ pretty=$(mktemp /tmp/flexible_arrays.data.sh > > <http://flexible_arrays.data.sh>.XXXXXX.c) > > > > echo -n "Flexible arrays accounting: " > > > > -for struct in $(pahole -F btf --sizes -- > > with_embedded_flexible_array $vmlinux | cut -f1) ; do > > +for struct in $(pahole -F btf --sizes -- > > with_embedded_flexible_array $vmlinux 2>/dev/null| cut -f1) ; do > > pahole $struct > $pretty > > > > # We need to check for just one tab before the comment as > > when expanding unnamed > > -- > > 2.43.5 > > ^ permalink raw reply [flat|nested] 5+ messages in thread
* [PATCH dwarves 2/2] tests/flexible_arrays: print flex array struct from same vmlinux 2024-10-24 11:26 [PATCH dwarves 0/2] flexible_arrays tests fixes Alan Maguire 2024-10-24 11:26 ` [PATCH dwarves 1/2] tests/flexible_arrays: redirect stderr to avoid warnings Alan Maguire @ 2024-10-24 11:26 ` Alan Maguire 1 sibling, 0 replies; 5+ messages in thread From: Alan Maguire @ 2024-10-24 11:26 UTC (permalink / raw) To: acme; +Cc: dwarves, Alan Maguire We get a list of flexible array structs from $vmlinux; use the same $vmlinux to print each struct. Signed-off-by: Alan Maguire <alan.maguire@oracle.com> --- tests/flexible_arrays.sh | 6 +++--- 1 file changed, 3 insertions(+), 3 deletions(-) diff --git a/tests/flexible_arrays.sh b/tests/flexible_arrays.sh index e9e3cb0..086db22 100755 --- a/tests/flexible_arrays.sh +++ b/tests/flexible_arrays.sh @@ -21,7 +21,7 @@ pretty=$(mktemp /tmp/flexible_arrays.data.sh.XXXXXX.c) echo -n "Flexible arrays accounting: " for struct in $(pahole -F btf --sizes --with_embedded_flexible_array $vmlinux 2>/dev/null| cut -f1) ; do - pahole $struct > $pretty + pahole $struct $vmlinux > $pretty # We need to check for just one tab before the comment as when expanding unnamed # structs with members with flexible arrays inside another struct we would mess @@ -54,14 +54,14 @@ for struct in $(pahole -F btf --sizes --with_embedded_flexible_array $vmlinux 2> if [ "$nr_embedded_flexible_arrays" != "$stat_nr_embedded_flexible_arrays" ] ; then test -n "$VERBOSE" && printf "struct %s: The number of embedded flexible arrays (%s) doesn't match the number of members marked as such (%s)\n" \ "$struct" "$stat_nr_embedded_flexible_arrays" "$nr_embedded_flexible_arrays" - test -n "$VERBOSE" && pahole $struct + test -n "$VERBOSE" && pahole $struct $vmlinux FAILED=1 fi if [ "$nr_flexible_arrays" != "$stat_nr_flexible_arrays" ] ; then test -n "$VERBOSE" && printf "struct %s: The number of flexible arrays (%s) doesn't match the number of members marked as such (%s)\n" \ "$struct" "$stat_nr_flexible_arrays" "$nr_flexible_arrays" - test -n "$VERBOSE" && pahole $struct + test -n "$VERBOSE" && pahole $struct $vmlinux FAILED=1 fi -- 2.43.5 ^ permalink raw reply related [flat|nested] 5+ messages in thread
end of thread, other threads:[~2024-10-24 13:45 UTC | newest]
Thread overview: 5+ messages (download: mbox.gz follow: Atom feed
-- links below jump to the message on this page --
2024-10-24 11:26 [PATCH dwarves 0/2] flexible_arrays tests fixes Alan Maguire
2024-10-24 11:26 ` [PATCH dwarves 1/2] tests/flexible_arrays: redirect stderr to avoid warnings Alan Maguire
[not found] ` <CA+JHD923mvOrr9JoKeVkeB1k2PTGSHsi7=6HPrV4tVNF1R=GWA@mail.gmail.com>
2024-10-24 13:01 ` Alan Maguire
2024-10-24 13:45 ` Arnaldo Carvalho de Melo
2024-10-24 11:26 ` [PATCH dwarves 2/2] tests/flexible_arrays: print flex array struct from same vmlinux Alan Maguire
This is a public inbox, see mirroring instructions for how to clone and mirror all data and code used for this inbox