On Wed, Dec 12, 2001 at 09:48:26AM +0100, Jean-Marc Lasgouttes wrote:
> John> no, each percentage is the percentage total number of samples
> John> registered against code found in that routine (I think libbfd
> John> attempts to locate inlined functions properly too though).
>
> Is that from op
> "John" == John Levon <[EMAIL PROTECTED]> writes:
John> On Tue, Dec 11, 2001 at 04:10:04PM +0100, Jean-Marc Lasgouttes
John> wrote:
>> How does one read this output? Do these include the time for the
>> childs?
John> no, each percentage is the percentage total number of samples
John> regist
On Tue, Dec 11, 2001 at 04:10:04PM +0100, Jean-Marc Lasgouttes wrote:
> How does one read this output? Do these include the time for the
> childs?
no, each percentage is the percentage total number of samples registered
against code found in that routine (I think libbfd attempts to locate inlin
> "John" == John Levon <[EMAIL PROTECTED]> writes:
John> On Sat, Dec 08, 2001 at 07:24:16PM +0100, Lars Gullik Bjønnes
John> wrote:
>> No, but with your profile it look as if it is the containers and
>> tools that create the overhead. With -O2 the case is completely
>> different, and "real" l
> "Herbert" == Herbert Voss <[EMAIL PROTECTED]> writes:
>> Would it be possible to see the book? I wonder what needs so much
>> convertion time... Would be a good candidate for profiling. Also,
>> were your two versions compiled with the same optimization level?
Herbert> I have no idea about
Jean-Marc Lasgouttes wrote:
>>"Herbert" == Herbert Voss <[EMAIL PROTECTED]> writes:
>>
>
> Herbert> a book of 298 pages written with 1.1.6 and a filesize of 1.9
> Herbert> MegaBytes - loading in 1.1.6 takes about 6 seconds loading
> Herbert> with converting into 1.2.0 takes more than 3
> "Herbert" == Herbert Voss <[EMAIL PROTECTED]> writes:
Herbert> a book of 298 pages written with 1.1.6 and a filesize of 1.9
Herbert> MegaBytes - loading in 1.1.6 takes about 6 seconds loading
Herbert> with converting into 1.2.0 takes more than 3 minutes!
Herbert> (Display Graphics was disab
> "John" == John Levon <[EMAIL PROTECTED]> writes:
John> On Sat, Dec 08, 2001 at 07:24:16PM +0100, Lars Gullik Bjønnes
John> wrote:
>> No, but with your profile it look as if it is the containers and
>> tools that create the overhead. With -O2 the case is completely
>> different, and "real" l
On Sat, 8 Dec 2001, Lars Gullik Bjønnes wrote:
> Herbert Voss <[EMAIL PROTECTED]> writes:
>
> | - in 1.1.6 there is often a construct like
> >
> |latex latex
> >
> |latex default
>
> it is the user that has put this contruct there.
Not me. I have seen this also when importing 1.1.6 docs
On Sat, Dec 08, 2001 at 07:24:16PM +0100, Lars Gullik Bjønnes wrote:
> No, but with your profile it look as if it is the containers and tools
> that create the overhead. With -O2 the case is completely different,
> and "real" lyx code begin to show up.
OK, that's a good point ...
I get :
Buffe
On Sat, Dec 08, 2001 at 06:50:42PM +0100, Lars Gullik Bjønnes wrote:
> | here is from my 100+ page document, mostly text + figures (not rendered).
> | It is fine ... perhaps it is a mathed thing :
>
> _and_ it makes little sense to do probiling on a binary that has not
> been compiled with full
John Levon wrote:
> On Sat, Dec 08, 2001 at 12:45:57PM +0100, Lars Gullik Bjønnes wrote:
>
>
>>| a book of 298 pages written with 1.1.6 and a filesize of 1.9 MegaBytes
>>
>
> lots of math ?
yes: math, floats, tables, ERT, special formatting, ...
Herbert
On Sat, Dec 08, 2001 at 05:38:28PM +0100, Lars Gullik Bjønnes wrote:
> | I think that this patch is wrong: The old code of insertErtContents is
> | correct, but the problem is that the wrong font is passed to it when the
> | method is called.
>
> I am not sure that I agree.
>
> | Consider the at
On Sat, Dec 08, 2001 at 12:45:57PM +0100, Lars Gullik Bjønnes wrote:
> | a book of 298 pages written with 1.1.6 and a filesize of 1.9 MegaBytes
lots of math ?
> I guess that we will need to run a profiler to find out where we use
> all that extra time. (both above cases)
here is from my 100+ p
Herbert Voss wrote:
>
>
> Dekel Tsur wrote:
>
> > On Sat, Dec 08, 2001 at 03:12:27PM +0100, Lars Gullik Bjønnes wrote:
> >
> >>Ok it seems that this patch solves the whole problem (buffer.C):
> >>@@ -388,12 +391,14 @@ bool Buffer::readLyXformat2(LyXLex & lex
> >> void Buffer::insertErtC
Dekel Tsur wrote:
> On Sat, Dec 08, 2001 at 03:12:27PM +0100, Lars Gullik Bjønnes wrote:
>
>>Ok it seems that this patch solves the whole problem (buffer.C):
>>@@ -388,12 +391,14 @@ bool Buffer::readLyXformat2(LyXLex & lex
>> void Buffer::insertErtContents(Paragraph * par, int & pos,
>>-
On Sat, Dec 08, 2001 at 03:12:27PM +0100, Lars Gullik Bjønnes wrote:
> Ok it seems that this patch solves the whole problem (buffer.C):
> @@ -388,12 +391,14 @@ bool Buffer::readLyXformat2(LyXLex & lex
> void Buffer::insertErtContents(Paragraph * par, int & pos,
> - Ly
Lars Gullik Bjønnes wrote:
> Ok it seems that this patch solves the whole problem (buffer.C):
>
> @@ -388,12 +391,14 @@ bool Buffer::readLyXformat2(LyXLex & lex
>
> #ifndef NO_COMPABILITY
> void Buffer::insertErtContents(Paragraph * par, int & pos,
> - LyXFont con
Lars Gullik Bjønnes wrote:
> Herbert Voss <[EMAIL PROTECTED]> writes:
>
> | - in 1.1.6 there is often a construct like
>
> |latex latex
>
> |latex default
>
> it is the user that has put this contruct there.
sure
>
> |which is converted into an empty ERT. it's not really
>
19 matches
Mail list logo
