59832 – sg3_utils fails to compile with USE=pic

Bug 59832 - sg3_utils fails to compile with USE=pic

Summary: sg3_utils fails to compile with USE=pic

Status:	RESOLVED FIXED

Alias:	None

Product:	Gentoo Linux
Classification:	Unclassified
Component:	New packages (show other bugs)
Hardware:	All All

Importance:	High major (vote)
Assignee:	Tim Yamin (RETIRED)

URL:
Whiteboard:
Keywords:

Depends on:
Blocks:

Reported:	2004-08-09 01:01 UTC by Cory Visi (RETIRED)
Modified:	2004-08-09 14:39 UTC (History)
CC List:	0 users

See Also:
Package list:
Runtime testing required:	---

Attachments
sg3_utils-llseek.patch (sg3_utils-llseek.patch,1.90 KB, patch) 2004-08-09 01:02 UTC, Cory Visi (RETIRED)	Details \| Diff
sg3_utils-1.06.ebuild.patch (sg3_utils-1.06.ebuild.patch,756 bytes, patch) 2004-08-09 01:02 UTC, Cory Visi (RETIRED)	Details \| Diff
View All Add an attachment (proposed patch, testcase, etc.)

Note You need to log in before you can comment on or make changes to this bug.

Description Cory Visi (RETIRED) gentoo-dev

2004-08-09 01:01:16 UTC

sg3_utils uses llseek.c from util-linux. solar fixed util-linux's llseek.c with files/util-linux-2.11z-pic.patch in July of 2003 (see the ChangeLog).

The _llseek() system call is actually the wrong way to call llseek now, and while his patch does compile with -fPIC, I believe my implementation is better (see Bug 51962).

With this change I believe we can compile safely with enhanced CFLAGS.

Finally, I removed the HAS_LONG_LONG check, since there is no configure script that will ever define this, and the rest of the code uses long long liberally.

Attached is an ebuild and source code patch.

I will submit the source code patch upstream, too.

Comment 1 Cory Visi (RETIRED) gentoo-dev

2004-08-09 01:02:17 UTC

Created attachment 37081 [details, diff]
sg3_utils-llseek.patch

Comment 2 Cory Visi (RETIRED) gentoo-dev

2004-08-09 01:02:41 UTC

Created attachment 37082 [details, diff]
sg3_utils-1.06.ebuild.patch

Comment 3 Tim Yamin (RETIRED) gentoo-dev

2004-08-09 14:39:24 UTC

Now fixed in CVS, it should reach the rsync mirrors shortly - closing as FIXED. Thanks for the patch and please reopen this bug if you have any further issues...