Bug#729187: qtbase-opensource-src: FTBFS on mips/mipsel
On Sat, Nov 09, 2013 at 10:22:25PM -0300, Lisandro Damián Nicanor Pérez Meyer wrote:
> tag 729187 moreinfo
> thanks
>
> On Sunday 10 November 2013 00:48:13 Aurelien Jarno wrote:
> > Package: qtbase-opensource-src
> > Version: 5.1.1+dfsg-5
> > Severity: important
> > Tags: patch upstream
> > Justification: fails to build from source
> [snip]
> > The patch below does exactly that. With it I was able to fully build Qt5
> > on a MIPS system. Would it be possible to include it in the next upload?
> > Thanks in advance.
>
> Hi Aurelien! First of all, thank *you* for doing this patch and explain it so
> clearly :)
>
> Allow me to make you some questions, which totally comes from my lack of
> knowledge wrt this.
>
> 1) Is it possible to surround this with something like #ifdef..#elseif..#blah?
> In other words, to detect if MIPS32 ISA or MIPS2 ISA is being used? If so, I
> would like to add those statements in order to have both options avilable and
> try to push the patch upstream. Or maybe I'm not understanding your
> explanation enough and the code itself works in all cases (but even then,
> maybe using defines gets faster code for MIPS32 ISA machines).
Yes, it is something possible, that said the code works in all case. It
basically just says to binutils "from now on use the MIPS32 instruction
set", and later "now switch back to the previous instruction set".
This means that the generated binary code is unchanged, just that it
compiles on more machines.
> 2) Supposing (1) above is possible: upstream wants patches trough Gerrit, so
> they know that people submitting them have agreed to the CLA. Would you mind
> to submit it to upstream's gerrit? In this way you will get the credit for it.
> Else I could do it on my self, but that doesn't feels correct at all.
>
I'll submit it to upstream gerrit, so we'll know if upstream consider it
as acceptable or not.
--
Aurelien Jarno GPG: 1024D/F1BCDB73
aurelien@aurel32.net http://www.aurel32.net
Reply to: