Re: USE_SHA1DC is broken in pu

2017-03-23 Thread Jonathan Nieder
Johannes Schindelin wrote: > On Wed, 22 Mar 2017, Jonathan Nieder wrote: > > Johannes Schindelin wrote: >>> As to the default of seriously slowing down all SHA-1 computations: >>> since you made that the default, at compile time, with no way to turn >>> on the faster computation, this will have a

Re: USE_SHA1DC is broken in pu

2017-03-23 Thread Junio C Hamano
Jeff King writes: > Side note: I also have a feeling that any operation that cares about > non-object sha1 performance is probably ripe for other, bigger > optimizations. If you update 300MB worth of index entries, then the > cost of computing a checksum over it isn't a

Re: USE_SHA1DC is broken in pu

2017-03-23 Thread Jeff King
On Thu, Mar 23, 2017 at 10:16:23AM -0700, Linus Torvalds wrote: > > If I write out an index, I should not suffer the slowdown from detecting > > collisions. > > The index case is a complete red herring. > > As already noted, the proper fix for the index case is to simply do it > asynchronously

Re: USE_SHA1DC is broken in pu

2017-03-23 Thread Linus Torvalds
On Thu, Mar 23, 2017 at 9:43 AM, Johannes Schindelin wrote: > > What I am saying is that this should be a more fine-grained, runtime knob. No it really shouldn't. > If I write out an index, I should not suffer the slowdown from detecting > collisions. The index case

Re: USE_SHA1DC is broken in pu

2017-03-23 Thread Johannes Schindelin
Hi Jonathan, On Wed, 22 Mar 2017, Jonathan Nieder wrote: > Johannes Schindelin wrote: > > > As to the default of seriously slowing down all SHA-1 computations: > > since you made that the default, at compile time, with no way to turn > > on the faster computation, this will have a major,

Re: USE_SHA1DC is broken in pu

2017-03-22 Thread Jonathan Nieder
Hi, Johannes Schindelin wrote: > As to the default of seriously slowing down all SHA-1 computations: since > you made that the default, at compile time, with no way to turn on the > faster computation, this will have a major, negative impact. Are you > really, really sure you want to do that? >

Re: USE_SHA1DC is broken in pu

2017-03-22 Thread Johannes Schindelin
Hi Junio, On Tue, 21 Mar 2017, Junio C Hamano wrote: > Johannes Schindelin writes: > > > On Fri, 17 Mar 2017, Lars Schneider wrote: > > > >> > On 17 Mar 2017, at 11:18, Lars Schneider > >> > wrote: > >> > > >> > Would it make sense/have

Re: USE_SHA1DC is broken in pu

2017-03-21 Thread Junio C Hamano
Johannes Schindelin writes: > On Fri, 17 Mar 2017, Lars Schneider wrote: > >> > On 17 Mar 2017, at 11:18, Lars Schneider >> > wrote: >> > >> > Would it make sense/have value to add a job to our TravisCI build [1] >> > that compiles Git in a

Re: USE_SHA1DC is broken in pu

2017-03-21 Thread Johannes Schindelin
Hi Lars, On Fri, 17 Mar 2017, Lars Schneider wrote: > > On 17 Mar 2017, at 11:18, Lars Schneider > > wrote: > > > > Would it make sense/have value to add a job to our TravisCI build [1] > > that compiles Git in a few variations with some high profile switches > > such

Re: USE_SHA1DC is broken in pu

2017-03-16 Thread Lars Schneider
> On 17 Mar 2017, at 11:18, Lars Schneider wrote: > > >> On 17 Mar 2017, at 03:22, Linus Torvalds >> wrote: >> >> I think there's a semantic merge error and it clashes with >> f18f816cb158 ("hash.h: move SHA-1 implementation selection

Re: USE_SHA1DC is broken in pu

2017-03-16 Thread Lars Schneider
> On 17 Mar 2017, at 03:22, Linus Torvalds > wrote: > > I think there's a semantic merge error and it clashes with > f18f816cb158 ("hash.h: move SHA-1 implementation selection into a > header file"). > > Suggested possible merge resolution attached. > >

Re: USE_SHA1DC is broken in pu

2017-03-16 Thread Linus Torvalds
On Thu, Mar 16, 2017 at 12:51 PM, Linus Torvalds wrote: > > I'll send a patch on top of 'next', which already has the header file changes. Patches sent. It all looked fairly straightforward to me, but maybe I missed something. Linus

Re: USE_SHA1DC is broken in pu

2017-03-16 Thread Linus Torvalds
On Thu, Mar 16, 2017 at 12:46 PM, Junio C Hamano wrote: > > That's easy to answer. What we have on 'pu' is a fair game for > wholesale replacement. That is the whole point of not merging > topics in flux to 'next' and declaring that 'pu' will constantly > rewind. Ok. I'll

Re: USE_SHA1DC is broken in pu

2017-03-16 Thread Junio C Hamano
Linus Torvalds writes: > I think there's a semantic merge error and it clashes with > f18f816cb158 ("hash.h: move SHA-1 implementation selection into a > header file"). > > Suggested possible merge resolution attached. > >Linus Obviously I have

Re: USE_SHA1DC is broken in pu

2017-03-16 Thread Junio C Hamano
Linus Torvalds writes: > On Thu, Mar 16, 2017 at 12:41 PM, Jeff King wrote: >> >> Potentially we should just eject sha1dc from "pu" for the moment. It >> needs re-rolled with the most recent version of the collision library >> (and I see Marc just

Re: USE_SHA1DC is broken in pu

2017-03-16 Thread Linus Torvalds
On Thu, Mar 16, 2017 at 12:41 PM, Jeff King wrote: > > Potentially we should just eject sha1dc from "pu" for the moment. It > needs re-rolled with the most recent version of the collision library > (and I see Marc just posted that they hit a stable point, which is > perhaps why

Re: USE_SHA1DC is broken in pu

2017-03-16 Thread Jeff King
On Thu, Mar 16, 2017 at 12:22:00PM -0700, Linus Torvalds wrote: > I think there's a semantic merge error and it clashes with > f18f816cb158 ("hash.h: move SHA-1 implementation selection into a > header file"). > > Suggested possible merge resolution attached. Yeah, your resolution makes sense.

USE_SHA1DC is broken in pu

2017-03-16 Thread Linus Torvalds
I think there's a semantic merge error and it clashes with f18f816cb158 ("hash.h: move SHA-1 implementation selection into a header file"). Suggested possible merge resolution attached. Linus Makefile | 2 +- hash.h | 2 ++ 2 files changed, 3 insertions(+), 1 deletion(-)