cml1.bbclass: Return sorted list of cfg files
PATCH 1/1 cml1.bbclass: Return sorted list of cfg files
Since these files are passed to merge_config.sh its better to use a sorted list, so we can be sure that this list is always fed in same order irrespective of python versions on host
--- meta/classes/cml1.bbclass | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)
meta/classes/cml1.bbclass +1 -1
@@ -6,7 +6,7 @@ def find_cfgs(d): if s.endswith('.cfg'): sources_list.append(s) - return sources_list + return sorted(sources_list) cml1_do_configure() { set -e
Signature · 1 line
2.31.1
Quoted 14 lines · Khem Raj
Since these files are passed to
merge_config.shits better to use a sorted list, so we can be sure that this list is always fed in same order irrespective of python versions on hostSigned-off-by Khem Raj <[email protected]>--- meta/classes/cml1.bbclass | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)meta/classes/cml1.bbclass +0
-0
meta/classes/cml1.bbclass def find_cfgs(d): · line 9
Quoted 5 lines
meta/classes/cml1.bbclass +1
-1@@ -6,7 +6,7 @@ def find_cfgs(d): if s.endswith('.cfg'): sources_list.append(s) - return sources_list + return sorted(sources_list)
Just as a thought, wouldn't it be better to have the ordering defined based on recipe/layer meta data?
Previously ordering was more or less based on the inode number of the files - now it's alphabetically.
I know that config fragments should be self containing, but seen enough real world examples where that isn't the case, so for them this might cause a breaking change (tbf the behavior before wasn't fully deterministic as well)
But this issue isn't limited to this particular piece of code here and might need some more general solution.
Quoted 8 lines
meta/classes/cml1.bbclass +0
-0cml1_do_configure() { set -e
Quoted 23 lines · Konrad Weihmann
Quoted 14 lines · Khem Raj
Since these files are passed to
merge_config.shits better to use a sorted list, so we can be sure that this list is always fed in same order irrespective of python versions on hostSigned-off-by Khem Raj <[email protected]>--- meta/classes/cml1.bbclass | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-)meta/classes/cml1.bbclass +0
-0meta/classes/cml1.bbclass def find_cfgs(d): · line 8
Quoted 4 lines
meta/classes/cml1.bbclass +1
-1@@ -6,7 +6,7 @@ def find_cfgs(d): if s.endswith('.cfg'): sources_list.append(s) - return sources_list + return sorted(sources_list)Just as a thought, wouldn't it be better to have the ordering defined based on recipe/layer meta data?
Quoted 7 lines
Previously ordering was more or less based on the inode number of the files - now it's alphabetically.
I know that config fragments should be self containing, but seen enough real world examples where that isn't the case, so for them this might cause a breaking change (tbf the behavior before wasn't fully deterministic as well)
Correct, config fragments can come from different places and it's quite common for latter fragments to amend some former fragments in the list. So, sorting them either way would sure enhance determinism, but surely break a lot of existing functionality out there...
Quoted 6 lines
But this issue isn't limited to this particular piece of code here and might need some more general solution.
Quoted 3 lines
meta/classes/cml1.bbclass +0
-0cml1_do_configure() { set -e
Signature · 4 lines
Regards, Denys Dmytriyenko <[email protected]> PGP: 0x420902729A92C964 - https://denix.org/0x420902729A92C964 Fingerprint: 25FC E4A5 8A72 2F69 1186 6D76 4209 0272 9A92 C964