Bug 430210 – Please stabilize net-misc/bfgminer-2.5.0 and x11-libs/amd-adl-sdk-4.0

• Status: RESOLVED FIXED
• Product: Gentoo Linux

[Description Luke-Jr 2012-08-06 18:08:37 UTC]

30 days in tree

[Comment 1 Anthony Basile 2012-08-06 18:09:39 UTC]

Arch teams, do your magic!

[Comment 2 Elijah "Armageddon" El Lazkani (amd64 AT) 2012-08-06 22:50:42 UTC]

Version does not exist (amd64)

[Comment 3 Anthony Basile 2012-08-06 23:47:48 UTC]

(In reply to comment #2)
> Version does not exist (amd64)

Ah yes!  Sorry, here's its history.  It was added to the tree July 6.  I removed it July 31 and Luke is now asking for it to be stabilized.

I reintroduced it.  Okay to restart stabilization?

[Comment 4 Agostino Sarubbo 2012-08-07 12:33:13 UTC]

(In reply to comment #3)
> I reintroduced it.  Okay to restart stabilization?
fine.

The following keyword changes are necessary to proceed:
#required by net-misc/bfgminer-2.5.0, required by =net-misc/bfgminer-2.5.0 (argument)
=dev-libs/jansson-2.3.1 ~amd64
#required by net-misc/bfgminer-2.5.0[adl], required by =net-misc/bfgminer-2.5.0 (argument)
=x11-libs/amd-adl-sdk-4.0 ~amd64

http://blogs.gentoo.org/ago/2012/07/06/repoman-check-before-file-stable-request/

[Comment 5 Anthony Basile 2012-08-07 12:36:05 UTC]

(In reply to comment #4)
> (In reply to comment #3)
> > I reintroduced it.  Okay to restart stabilization?
> fine.
> 
> The following keyword changes are necessary to proceed:
> #required by net-misc/bfgminer-2.5.0, required by =net-misc/bfgminer-2.5.0
> (argument)
> =dev-libs/jansson-2.3.1 ~amd64
> #required by net-misc/bfgminer-2.5.0[adl], required by
> =net-misc/bfgminer-2.5.0 (argument)
> =x11-libs/amd-adl-sdk-4.0 ~amd64
> 

The request comes with the understanding that the deps should be stabilized too.

[Comment 6 Anthony Basile 2012-08-09 15:57:12 UTC]

> > =x11-libs/amd-adl-sdk-4.0 ~amd64

Luke and I maintain x11-libs/amd-adl-sdk-4.0.  It has been in the tree since May 22 no bugs.  Its okay to stabilize for bfgminer.

[Comment 7 Agostino Sarubbo 2012-08-11 19:47:47 UTC]

(In reply to comment #6)
> > > =x11-libs/amd-adl-sdk-4.0 ~amd64
> 
> Luke and I maintain x11-libs/amd-adl-sdk-4.0.  It has been in the tree since
> May 22 no bugs.  Its okay to stabilize for bfgminer.

Then please edit the summary next time.

[Comment 8 Agostino Sarubbo 2012-08-11 20:35:36 UTC]

amd64 stable

[Comment 9 Johannes Huber (RETIRED) 2012-08-22 21:26:46 UTC]

The following keyword changes are necessary to proceed:
#required by virtual/opencl-0-r2, required by net-misc/bfgminer-2.5.0, required by =net-misc/bfgminer-2.5.0 (argument)
=dev-util/intel-ocl-sdk-1.5 **

The following license changes are necessary to proceed:
#required by virtual/opencl-0-r2, required by net-misc/bfgminer-2.5.0, required by =net-misc/bfgminer-2.5.0 (argument)
>=dev-util/intel-ocl-sdk-1.5 Intel-SDP

[Comment 10 Luke-Jr 2012-08-22 23:24:48 UTC]

All OpenCL options need to be keyworded first, even though BFGMiner doesn't require OpenCL at all? How does the license for it come into the picture?

[Comment 11 Anthony Basile 2012-09-15 01:38:10 UTC]

ping x86

[Comment 12 Paweł Hajdan, Jr. (RETIRED) 2012-10-15 01:08:29 UTC]

(In reply to comment #11)
> ping x86

Please see https://bugs.gentoo.org/show_bug.cgi?id=430210#c9 . This is what prevents me from testing this.

[Comment 13 Luke-Jr 2012-10-15 01:18:48 UTC]

(In reply to comment #12)
> (In reply to comment #11)
> > ping x86
> 
> Please see https://bugs.gentoo.org/show_bug.cgi?id=430210#c9 . This is what
> prevents me from testing this.

That's a runtime dependency for USE=opencl, and probably a bug itself since Intel's OCL SDK won't satisfy the dependency (but any non-CPU OpenCL SDK will). You should be able to test CPU mining just fine without it (USE=-opencl). Even if you could test with OpenCL, you couldn't test any of the FPGA drivers without that specific hardware, so I don't think it'd be reasonable to consider OpenCL testing necessary to stabilization (as long as 3rd parties can confirm it works, which it does).

[Comment 14 Paweł Hajdan, Jr. (RETIRED) 2012-10-16 03:42:18 UTC]

(In reply to comment #13)
> so I don't think it'd be
> reasonable to consider OpenCL testing necessary to stabilization (as long as
> 3rd parties can confirm it works, which it does).

That's not the problem. You can't stabilize package with unstable dependencies.

[Comment 15 Luke-Jr 2012-10-16 03:45:57 UTC]

I guess that's why blueness added amd-adl-sdk to the stablereq; it satisfies the same dependency.

[Comment 16 Luke-Jr 2012-10-16 03:47:12 UTC]

Sorry, getting late here... amd-adl-sdk does NOT satisfy that dependency; ati-drivers does, and is already stable.

[Comment 17 Agostino Sarubbo 2013-02-25 13:30:56 UTC]

x86 stable. Last arch, closing