search for: synthetically

On Fri, Dec 15, 2017 at 11:13 AM, Sean Silva <chisophugis at gmail.com> wrote: > > > On Fri, Dec 15, 2017 at 10:22 AM, Easwaran Raman <eraman at google.com> > wrote: > >> >> >> On Fri, Dec 15, 2017 at 12:22 AM, Sean Silva <chisophugis at gmail.com> >> wrote: >> >>> IIUC, this proposal is just saying that we should infer a

On Fri, Dec 15, 2017 at 12:22 AM, Sean Silva <chisophugis at gmail.com> wrote: > IIUC, this proposal is just saying that we should infer a static profile > for entry counts just like we do for branch probabilities. In the case of > entry counts, we do not hide that information behind an analysis like BPI, > so currently just annotating synthetic PGO entry counts is a simple >

On Fri, Dec 15, 2017 at 11:56 AM, Sean Silva <chisophugis at gmail.com> wrote: > > > On Fri, Dec 15, 2017 at 11:27 AM, Xinliang David Li <davidxl at google.com> > wrote: > >> >> >> On Fri, Dec 15, 2017 at 11:13 AM, Sean Silva <chisophugis at gmail.com> >> wrote: >> >>> >>> >>> On Fri, Dec 15, 2017 at 10:22

Functions in LLVM IR have a function_entry_count metadata that is attached in PGO compilation. By using the entry count together with the block frequency info, the compiler computes the profile count of call instructions based on which the hotness/coldness of callsites can be determined. Experiments have shown that using a higher threshold for hot callsites results in improved runtime performance

My $0.02: LB is baked in to nut behavior. So if a UPS doesn't report LB, then it makes sense to synthesize it. Synthetic LB is the cleanest fix at the earlier processing point. One could have a config that is basically synthetic-lowbatt-percent that generates LB if (as Jim says) on battery, not calibrating, and batt% <= that. One could have a similar config

Sure. One thing I want to remind you is that there is a place in Writer.cpp where we assume all synthetic sections were appended to end of Sections list. Look for `llvm::reverse`. If you add synthetic section thunks right before/after non-synthetic sections, you also want to change that. On Tue, Nov 29, 2016 at 1:18 PM, Simon Atanasyan <simon at atanasyan.com> wrote: > Hi, > >

Hi, I've prototyped a pair of passes which 1) add synthetic debug info to a Module and 2) determine how much of that info is lost. The idea is to make it easier to write targeted test cases for debug info preservation. For example, here is a quick way to test that Mem2Reg doesn't drop debug info from one test input: > ; RUN: opt < %s -debugify -mem2reg -check-debugify -S

Hi, Sorry for delay with reply. It looks like now thunks can be implemented as a synthetic sections. In that case we give flexible solution and will be able to put thunks before/after related sections, using different alignment etc. As far as I know BFD linker uses the same approach at least for MIPS thunks. I will try to implement this idea. On Thu, Sep 8, 2016 at 1:44 AM, Rui Ueyama <ruiu

I used Microsoft text to speech with speex. It sounded reasonably well. I didn't experiment with various quality settings (8KHz, default quality). I was capturing audio from a sound card, not the direct wave data from the TTS engine. .a -----Original Message----- From: speex-dev-bounces@xiph.org [mailto:speex-dev-bounces@xiph.org] On Behalf Of Reed Hedges Sent: Friday, May 26, 2006 7:23 AM

Hi, Merry Christmas to all. :) I have been trying out new 1.2beta3 here and facing some noise problems, (I can't say this is 1.2beta3 problem because I haven't tried out previous version yet) and was wondering if anyone also has faced and solved similar difficulties. When the speex encoding/decoding (wideband) process starts, the voice is crispy clear with almost no noise in the

I was hoping to use mi to generate a synthetic version of a database. The strategy (see code below) was simple: use the diamonds dataset from ggplot2, subset it focus on 3K single-color, then create a blank record for every "real" record, and throw the new dataset at mi to see if it would populate the blank records. I kept getting failed convergence. I think I have simplified the

For some queries it is complicated to express them directly in SQL. Even more so when they are not hand-written, but generated from something the user entered. Say, the user wants to search for person records where the name contains some string. Now, while the Rails model may have a name attribute, the database table does not. There the name is decomposed into title, firstname, lastname and

On Tue, Nov 28, 2017 at 11:27 AM Reid Kleckner <rnk at google.com> wrote: > On Tue, Nov 28, 2017 at 11:23 AM, David Blaikie via llvm-dev < > llvm-dev at lists.llvm.org> wrote: > >> Alternatively we can really say this header is a textual header - it's >> included generally only once in a whole program, the functions are called >> only once, etc. Though

Good Morning I want to perform a synthetic control method with R. For this purpose, I created the following code: # Re-load packages library(Synth) library(readxl) # Pfadeinstellung Excel-Blatt excel_file_path <- ("C:\\Users\\xxxxx\\Desktop\\DATA_INVESTMENTVOLUMEN_FOR_R_WITHOUT_NA.xlsx") # Load the Excel file INVESTMENTVOLUME <- read_excel(excel_file_path) #

Hi list, We have an ancient hack that fuse not just passes on the statvfs data it's getting from the storage, but tweaks it by setting f_bsize / f_frsize to values of its own preference. [1] The supposed advantage is that f_bsize serves as a hint to applications for the preferred io size. (And regarding f_frsize -- in Linux it's a historical workaround for certain bugs in userspace[2]

(I'm not subscribed, please keep me CC'd) Hi, since a few months I'm experiencing repeated dovecot crashes. I mostly see them in the logs, I haven't had any users complain, so I haven't looked closely until now. I was originally using an older version fro mDebian and hoped that a newer future version would already fix this, but I just upgraded to version 1:2.2.32-2 from

On Tue, Nov 28, 2017 at 11:23 AM, David Blaikie via llvm-dev < llvm-dev at lists.llvm.org> wrote: > Alternatively we can really say this header is a textual header - it's > included generally only once in a whole program, the functions are called > only once, etc. Though that does seem a little unfortunate on principle but > not much practical problem with it, I think.

Carsten Aulbert <carsten.aulbert at aei.mpg.de> writes: > Hi all, > > On 5/19/23 15:11, Greg Troxel wrote: >> LB is baked in to nut behavior. So if a UPS doesn't report LB, then >> it makes sense to synthesize it. Synthetic LB is the cleanest fix at >> the earlier processing point. > > I fully agree and thus this ought to be done in/near uspd

On 09/10/2015 17:53, Roman Kagan wrote: > > I really don't like this auto-EOI extension, but I guess that's the > > spec. :( If it wasn't for it, you could do everything very easily in > > userspace using Google's proposed MSR exit. > I guess you're right. We'd probably have to (ab)use MSI for SINT > delivery, though. Not really an issue, as MSI on

Am 09.10.2015 um 16:42 schrieb Paolo Bonzini: > Christian, the question for you is towards the end... [....] > >> --- a/virt/kvm/irqchip.c >> +++ b/virt/kvm/irqchip.c >> @@ -144,11 +144,13 @@ static int setup_routing_entry(struct kvm_irq_routing_table *rt, >> >> /* >> * Do not allow GSI to be mapped to the same irqchip more than once. >> -