reentrant functions

Bernhard Reutner-Fischer rep.dot.nop at gmail.com
Thu Sep 11 14:22:55 UTC 2008


On Tue, Sep 09, 2008 at 05:41:40PM +0200, Bernhard Reutner-Fischer wrote:
>On Tue, Sep 09, 2008 at 12:45:39PM +0200, Bernhard Reutner-Fischer wrote:
>>On Sun, Jun 08, 2008 at 02:18:03PM +0200, Bernd Schmidt wrote:
>>> Bernhard Fischer wrote:
>>>> Just to make sure i understand correctly:
>>>>
>>>> - If a reentrant function exists then you want the non-reentrant func to
>>>>   be optional (via a central knob).
>>>> - If non-reentrant funcs are off then they are aliased to their
>>>>   reentrant counterpart.
>>>
>>> No, they have different calling conventions, so they'd just not exist.


>What do people think about scandir(). Would you prefer to use readdir_r
>for it or should it use a not publically visible __libc_readdir(DIR*)?
>
>Is it worthwile to put the definition of TTYNAME_BUFLEN into a private
>ttyname.h or will everybody remember to keep the two in sync?
>Alternatively it could come from the buildsystem, i have no preference.
>
>An unrelated hunk in the patch* above begs the question as to whether or
>not __libc_sigaction() should use the rt_sigaction syscall internally
>if the user turned off (the external API) UCLIBC_HAS_REALTIME.
>rt_sigaction is currently used internally and back then i thought and
>still think this is ok.  Opinions?

Perhaps somebody has time to test the attached patch. The final version
will make sure not to add new relocations.
-------------- next part --------------
A non-text attachment was scrubbed...
Name: optional_non-reentrant.01.patch
Type: text/x-diff
Size: 31890 bytes
Desc: not available
Url : http://lists.busybox.net/pipermail/uclibc/attachments/20080911/b68564e4/attachment.bin 


More information about the uClibc mailing list