Re: [PATCH] decodetree: Do not remove output_file from /dev

[Richard Henderson <richard.henderson@linaro.org>]

On 5/26/23 12:52, Peter Maydell wrote:
> On Fri, 26 May 2023 at 18:40, Richard Henderson
> <richard.henderson@linaro.org> wrote:
>>
>> Nor report any PermissionError on remove.
>>
>> Previously we were testing with "> /dev/null", but that's not easy
>> to do with meson test(), so we want to use '-o /dev/null' instead.
>> That works fine for all of the existing tests, where all errors are
>> diagnosed before opening the output file.  However, PMM's named field
>> patch set diagnoses cycle errors during output.  This is fair, but
>> we need to be more careful with the remove.
>>
>> Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
>> ---
>>   scripts/decodetree.py | 7 ++++++-
>>   1 file changed, 6 insertions(+), 1 deletion(-)
>>
>> diff --git a/scripts/decodetree.py b/scripts/decodetree.py
>> index e4ef0a03cc..a9a0cd0fa3 100644
>> --- a/scripts/decodetree.py
>> +++ b/scripts/decodetree.py
>> @@ -71,7 +71,12 @@ def error_with_file(file, lineno, *args):
>>
>>       if output_file and output_fd:
>>           output_fd.close()
>> -        os.remove(output_file)
>> +        # Do not try to remove e.g. -o /dev/null
>> +        if not output_file.startswith("/dev"):
>> +            try:
>> +                os.remove(output_file)
>> +            except PermissionError:
>> +                pass
> 
> Maybe rather than hardcoding /dev, only try to delete the file
> if it's a normal file, i.e.:
>         if os.path.isfile(output_file):
>             os.remove(output_file)

Good idea.


r~