search for: _once_

For current FunctionPass passes, each function in the current module will be visited _once_ under some order. If I have an algorithm that needs to visit each function more than once in order to generate the final result, how could it be arranged under the current LLVM Pass framework? Could somebody point me a pass that currently doing this? Thank you Chuck

Chuck Zhao wrote: > For current FunctionPass passes, each function in the current module > will be visited _once_ under some order. > > If I have an algorithm that needs to visit each function more than once > in order to generate the final result, how could it be arranged under > the current LLVM Pass framework? > > Could somebody point me a pass that currently doing this? > You may...

...trix. And BTW, what a fantastic package. My problem is that I have to carry out repeated Cholesky factorization of a spares symmetric matrices, say Q_1, Q_2, ...,Q_n, where the Q's have the same non-zero pattern. I know in this case one does only need to carry out the symbolic factorization _once_ and then follow that up with a numerical factorization for each of the Q_i's (re-using the general symbolic factorization each time). Does anybody know if this is possible using the Matrix package? Thanks, Gardar Johannesson Lawrence Livermore National Laboratory

Description: 'lapply' returns a list of the same length as 'X', each element of which is the result of applying 'FUN' to the corresponding element of 'X'. I expect that when I do > lapply(vec,f) f would be called _once_ for each component of vec. this is not what I see: parse.num <- function (s) { cat("parse.num1\n"); str(s) s <- as.character(s) cat("parse.num2\n"); str(s) if (s == "N/A") return(s); as.numeric(gsub("M$","e6",gsub("B$",&q...

...don't understand what you meant here. Let me explain (even though I propose taking all the LFSR stuff out). pv_lock_hash_bit is a runtime variable, therefore it cannot compile time evaluate the forest of if statements required to compute the taps value. Therefore its best to compute the taps _once_, and this seems like a good place to do so. > >>+ goto done; > >>+ } > >>+ } > >>+ > >>+ hash = lfsr(hash, pv_lock_hash_bits, 0); > >Since pv_lock_hash_bits is a variable, you end up running through that > >massive if() forest to find...

...don't understand what you meant here. Let me explain (even though I propose taking all the LFSR stuff out). pv_lock_hash_bit is a runtime variable, therefore it cannot compile time evaluate the forest of if statements required to compute the taps value. Therefore its best to compute the taps _once_, and this seems like a good place to do so. > >>+ goto done; > >>+ } > >>+ } > >>+ > >>+ hash = lfsr(hash, pv_lock_hash_bits, 0); > >Since pv_lock_hash_bits is a variable, you end up running through that > >massive if() forest to find...

...o blab as much as you want and argue with > people over tangent issues proving you are right, Dude, at this point, I'm tired of having my experience cut, sliced and diced by people who are like, "WTF does he think he is?" So you're right, I'm going to just state something _once_ and if people don't want to believe it, I'm just going to leave it. I tried several times in that thread, but yes, you are correct, I didn't stop, and I kept posting too. My apologies. > or you get to maintain "non-troll" status. Pick one, but don't > tangentiall...

...; > Here, the entry for cpu3 will be removed leading to panic when > pv_hash_find() can find the entry. So the hash entry can only be > removed if the other cpu has no chance to see _Q_SLOW_VAL. Still confused. Afaict that cannot happen with my code. We only do the cmpxchg(, _Q_SLOW_VAL) _once_. Only on the first time around that loop will we hash the lock and set the slow flag. And cpu3 cannot come in on the same entry because we've not yet released the lock when we find and unhash.

...; > Here, the entry for cpu3 will be removed leading to panic when > pv_hash_find() can find the entry. So the hash entry can only be > removed if the other cpu has no chance to see _Q_SLOW_VAL. Still confused. Afaict that cannot happen with my code. We only do the cmpxchg(, _Q_SLOW_VAL) _once_. Only on the first time around that loop will we hash the lock and set the slow flag. And cpu3 cannot come in on the same entry because we've not yet released the lock when we find and unhash.

Markersdown: I was fooling around with lua and decided to write a markdown parser using the terrific lpeg library. Here's the result: http://github.com/jgm/lunamark/tree/master There are already two markdown libraries for lua, one a native lua implementation based on global substitutions, the other a binding to discount. What makes lunamark different is that it is based on a PEG grammar

...ery grateful if anyone could tell me what option to use. If not, I think I'll try and make a patch. Cheers Jorge P.S.: I have just enabled and I'm trying "security = domain" with a great, great success. All conections go _really_ faster now, despite the fact that users are asked _ONCE_ for the password, and no more. Keep up the good work. -- Jorge Gonzalez <jorgegv@icai.upco.es> -o) ICAI - Universidad Pontificia Comillas /\\ Administrador de Sistemas _\_v

...that one without a protocol change. It has been suggested that because we can't fix _that_ bug, we shouldn't fix _this_ bug either, for 'consistency'. That logic seems a little flawed to me. Besides, I've tripped over the 'wrong display' thing _many_ times and I have not _once_ been bothered by the fact that I'm using the 'wrong' agent.) -- Configure bugmail: http://bugzilla.mindrot.org/userprefs.cgi?tab=email ------- You are receiving this mail because: ------- You are watching the assignee of the bug.

Hi, [Context : ubuntu 11.11 64bits, cman, clvmd, gfs2, ctdb, samba, likewise-open, all running fine except...] I've setup ctdb to manage public_addresses (to manage one virtual ip actually), and I've explicitely told smb.conf that netbios name = foobar_cluster I've done that *after* I've manage to make samba, likewise-open and AD to work nicely together. I guess now all the

I have moved the R-announce mailing list from the Majordomo software to a combination of procmail & mailman an hour or so ago. You find a link in the footer of this message giving an URL where you can modify your subscription, look at the archives (everything from the beginning in 1997) and more. CHANGE: Postings to R-announce are now moderated, ie. they have to be approved by a list

I have moved the R-announce mailing list from the Majordomo software to a combination of procmail & mailman an hour or so ago. You find a link in the footer of this message giving an URL where you can modify your subscription, look at the archives (everything from the beginning in 1997) and more. CHANGE: Postings to R-announce are now moderated, ie. they have to be approved by a list

On Mon, Apr 06, 2015 at 10:55:44PM -0400, Waiman Long wrote: > +++ b/kernel/locking/qspinlock_paravirt.h > @@ -0,0 +1,321 @@ > +#ifndef _GEN_PV_LOCK_SLOWPATH > +#error "do not include this file" > +#endif > + > +/* > + * Implement paravirt qspinlocks; the general idea is to halt the vcpus instead > + * of spinning them. > + * > + * This relies on the

On Mon, Apr 06, 2015 at 10:55:44PM -0400, Waiman Long wrote: > +++ b/kernel/locking/qspinlock_paravirt.h > @@ -0,0 +1,321 @@ > +#ifndef _GEN_PV_LOCK_SLOWPATH > +#error "do not include this file" > +#endif > + > +/* > + * Implement paravirt qspinlocks; the general idea is to halt the vcpus instead > + * of spinning them. > + * > + * This relies on the

...d catalog for puppetmaster in 722.00 seconds Sun May 18 10:55:06 +0200 2008 Puppet (notice): Compiled catalog for puppetmaster in 691.28 seconds Sun May 18 11:17:32 +0200 2008 Puppet (notice): Caught TERM; shutting down So if I start the puppetd --test (after a previous restart of the master) only _once_, I can see the following behaviour: 1. puppetmaster''s memory is growing very fast and very quick, up to a not yet known level. 2. client (problemhost in the logs) exits with a gateway timeout, which can''t be adjusted by reconfiguring nginx. (or did I miss anything? maybe a firewa...

Hi Abby, Thanks a lot for your paraphrasing and your suggestion! The problem of wrapping the list into a S3/S4 object, i.e. subclassing array or matrix, is that one also has to define a bunch of methods for subsetting, joining, etc, in order to make it behave like a list array. The reason is that most R functions for subsetting, joining, etc. do not preserve class attributes of the input, which

Hi , I have set up a 2-node CTDB cluster serving NFS and CIFS authenticating Windows and Linux users via Active Directory. The setup works fine, except only one server in the CTDB-cluster is able to join the AD domain at a given instance. If you manually add the other server into AD, the already connected server gets disconnected. There is no specific error message logged in /var/log/message or