| 1 |
On Wed, May 02, 2012 at 11:23:48AM -0400, Matt Turner wrote: |
| 2 |
> Jim? |
| 3 |
|
| 4 |
Sorry, I am behind on my gentoo-dev reading, thanks Matt for bringing |
| 5 |
this thread to my attention. Here are the two main concerns I've |
| 6 |
gathered from the rest of the thread: |
| 7 |
|
| 8 |
On Wed, Apr 25, 2012 at 11:04:08PM -0600, Ryan Hill wrote: |
| 9 |
> Arg, no. Please just print the warning if the host doesn't do SSE2. |
| 10 |
> There's no reason to have a USE flag here (and _really_ no reason to |
| 11 |
> make it fatal), especially for an instruction set that every system |
| 12 |
> has supported for over a decade. |
| 13 |
|
| 14 |
It's true that probably "most" people have sse2. And this change won't |
| 15 |
affect all those people that have sse2. It only affects people that |
| 16 |
don't have sse2. |
| 17 |
|
| 18 |
The run-time failure case (flash just doesn't work) is tricky to |
| 19 |
diagnose, so I want the user to be aware of this at install time. Thus |
| 20 |
I feel the check should (1) be fatal and (2) provide instructions that |
| 21 |
10.3 can still be used. |
| 22 |
|
| 23 |
The reason there's a USE flag is so that binary packagers can forcibly |
| 24 |
skip the 'die' if they know they're building for an sse2 SYSTEM. This |
| 25 |
is only needed if you're running on a machine without sse2, packaging for |
| 26 |
a machine with sse2. |
| 27 |
|
| 28 |
On Thu, Apr 26, 2012 at 05:06:45PM -0300, Alexis Ballier wrote: |
| 29 |
> wouldnt adding a sse2 useflag and putting it in REQUIRED_USE solve the |
| 30 |
> problem ? |
| 31 |
> |
| 32 |
> afaik portage wont even try to upgrade if people have -sse2 |
| 33 |
|
| 34 |
This is an interesting idea, as it could automatically downgrade users to |
| 35 |
10.3 without requiring any masking. I like that. |
| 36 |
|
| 37 |
However, the problem I have with this approach is that it doesn't |
| 38 |
address people who do have sse2 instructions available but have not set |
| 39 |
the USE flag. I noticed that our amd64 profile does set USE=sse2 |
| 40 |
automatically. That's good. But x86 users would still have to set this |
| 41 |
manually. And people without the flag will be forcibly downgraded to |
| 42 |
10.3 for no good reason. And as Ryan accurately points out, most people |
| 43 |
*do* have sse2 instructions available. |
| 44 |
|
| 45 |
I prefer checking /proc/cpuinfo because that way I'm *sure* that the |
| 46 |
user's system has the required instruction set. |
| 47 |
|
| 48 |
The only way I think that this decision can really be made would be to |
| 49 |
have some sort of usage statistics on how many x86 setups there are |
| 50 |
where the CPU has sse2 available but the user has not set USE=sse2. Any |
| 51 |
idea how we can know how many this is? |
| 52 |
|
| 53 |
Hmmm... I suppose the other thing I could do is put the /proc/cpuinfo |
| 54 |
check into the 10.3 ebuild, and warn users if they're installing 10.3 |
| 55 |
and *should* be setting USE=sse2 to get 11.2 instead.... Maybe that's the |
| 56 |
best solution! |
| 57 |
|
| 58 |
Thanks for the suggestions, I'll try to get at this change soon. |
| 59 |
|
| 60 |
-- |
| 61 |
Jim Ramsay |
Archives