To comment on the following update, log in, then open the issue:
http://www.openoffice.org/issues/show_bug.cgi?id=108371
User sb changed the following:
What |Old value |New value
================================================================================
Ever confirmed| |1
--------------------------------------------------------------------------------
Status|UNCONFIRMED |STARTED
--------------------------------------------------------------------------------
------- Additional comments from [email protected] Tue Jan 26 16:37:19 +0000
2010 -------
@ecatmur: Thanks for the pointer. Not sure if the apparent non-existence of a
Linux x86 ABI should make me laugh or cry. Anyway, as it reportedly helps, we
can integrate the patch. Just a few questions:
>From the placement of the esp adjustments, I gather that the requirement is
>that
esp is 16 byte aligned just prior to a call (i.e., the return address always
goes to an 0x...C address); right?
Then, in call.s, privateSnippetExecutorVoid, doing 5*pushl = 20 byte, plus 4
byte return address, would need the esp adjustment, too, right?
Also, in all the other adjustments in call.s, instead of doing something like
subl $0x4,%esp
andl $0xfffffff0,%esp
one could instead exploit the knowledge that upon entry to the function esp =
0x...C and do
subl $0x8,%esp
instead (for an original subtract of $0x4, and a total of 5 pushl) to ensrure
that afterwards exp = 0x...0. Right? (Which would save an instruction, and
would waste less stack space on average for environments that only do 4 byte
alignment).
---------------------------------------------------------------------
Please do not reply to this automatically generated notification from
Issue Tracker. Please log onto the website and enter your comments.
http://qa.openoffice.org/issue_handling/project_issues.html#notification
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
---------------------------------------------------------------------
To unsubscribe, e-mail: [email protected]
For additional commands, e-mail: [email protected]
