RFR(M): 8020775: PPC64 (part 12): posix signal printing
coleen.phillimore at oracle.com
Fri Aug 16 10:47:20 PDT 2013
On 8/16/2013 1:12 PM, Vladimir Kozlov wrote:
> On 7/28/13 7:46 PM, David Holmes wrote:
>> On 26/07/2013 9:11 AM, Lindenmaier, Goetz wrote:
>>> we'd like to contribute our posix signal printing.
>> This really isn't "posix signals" as it combines all the OS specific
>> signal definitions into one chunk of code.
> Currently it is the only place in our sources where we can combine
> code for all unix variations.
Yes, we're renaming "posix" to "xnix", so it's good to not duplicate the
code in all the unix variations with this change, just because it's not
>> + // A number high enough to contain all possible signal numbers.
>> + #define MAX_SIGNAL_NUMBER 70
>> Why do you need this when you use sigaddset to check validity anyway?
>> What is this "maximum signal number" meant to represent anyway? The
>> maximum signal number on the platform, or the maximum signal number for
>> a signal that the JVM will install a handler for?
>> The runtime team will need to take a good look at this. Personally I'd
>> rather not see all the different OS stuff piled in together. I'd
>> certainly like to see as little duplication as possible, but I'd rather
>> platform specific stuff was dealt with in platform specific files.
> It is just printing unification to avoid a lot of duplication.
> I agree that methods is_valid_signal() could stay in platform specific
> files. But to have the rest, signals names and print methods, in one
> place I think is good.
>>> We implemented some routines to print signal and sa_flags information
>>> in the os/posix files, and call them from
>>> os::print_siginfo and print_signal_handler() in the various unix
>>> variant directories.
>>> The output is a bit more verbose than the existing version.
>>> We contribute this here, as our aix code uses this too.
>>> Please review this and test it if you think we should add this.
>>> We'd appreciate it.
>>> Thanks and best regards,
More information about the hotspot-dev