16114 - does xptcall code for Solaris x86 work right?

Status: VERIFIED FIXED

(Core :: XPConnect, defect, P3)

Description

[John Bandhauer]

I saw the posting in the patches group about using the the unixish-x86 xptcall
code for Solaris x86. Does this work right? Pass tests? etc?

It reminded me that Arthur Jones <[e-mail]> had sent me an xptcall
implementation for Solaris x86 some time back. I replied to him at the time that
it needed some changes (at least remove the printf from this low-level code).
I'm sad to admit that I forgot about this.

Are there any bits of his implementation that the Solaris x86 needs which differ
from the more generic unixish-x86 implementation? I'm not setup to test this and
don't have any useful knowledge of this platform.

I'll attach Arthur's files to this bug.

Comment 1

[John Bandhauer]

Attachment: (1 of 3) files from Arthur Jones

Comment 2

[John Bandhauer]

Attachment: (2 of 3) files from Arthur Jones

Comment 3

[John Bandhauer]

Attachment: (3 of 3) files from Arthur Jones

Comment 4

[John Bandhauer]

Let's dump this on shaver and see if he squirms.

Comment 5

[Mike Shaver (:shaver -- probably not reading bugmail closely)]

I want to do something with this bug, but I don't exactly know what.  Arthur, do
you have any thoughts on jband's questions?  If the xptcall stuff works for you,
I'm happy to check it in.

Comment 6

[leger]

Moving all XPConnect QA contact to rginda

Comment 7

[Mike Shaver (:shaver -- probably not reading bugmail closely)]

Arthur?  Do you want this stuff to go in?

Comment 8

[Rich Burridge]

Now that the xptcall assembler code works for Solaris Intel compiled with
eith the Gnu compilers or the Sun native compilers, can this bug be closed out?

Comment 9

[Donnie Cranford]

Concurring that xptcall is working on SolarisX86 

Comment 10

[Mike Shaver (:shaver -- probably not reading bugmail closely)]

Sold!

Comment 11

[Rich Burridge]

Verified.