An email thread started from a question from Leo [1] stimulated
me to add a QQz on Paul's experience back in 2009.
As \QuickQuizLabel{} inside \QuickQuizSeries{} doesn't work in
-nq builds, define \QuickQuizLabelRel{}{} and put it in front
of the series.
Link: [1] https://www.spinics.net/lists/perfbook/msg03824.html
Signed-off-by: Akira Yokosawa <akiyks@xxxxxxxxx>
Cc: Leonardo Brás <leobras.c@xxxxxxxxx>
---
v2: Place new QQz next to QQz 3.7
--
cpu/overheads.tex | 28 +++++++++++++++++++++++-----
qqz.sty | 16 ++++++++++++++++
2 files changed, 39 insertions(+), 5 deletions(-)
diff --git a/cpu/overheads.tex b/cpu/overheads.tex
index 0d8270bf6e17..a89c71158bf9 100644
--- a/cpu/overheads.tex
+++ b/cpu/overheads.tex
@@ -290,12 +290,15 @@ nanoseconds, or more than seven hundred clock cycles.
A CAS operation consumes almost a full microsecond, or almost two
thousand clock cycles.
-\QuickQuiz{
+\QuickQuizLabelRel{\QspeedOfLightAtoms}{1} % cann't put label inside QQSeries
+
+\QuickQuizSeries{%
+\QuickQuizB{
Surely the hardware designers could be persuaded to improve
this situation!
Why have they been content with such abysmal performance
for these single-instruction operations?
-}\QuickQuizAnswer{
+}\QuickQuizAnswerB{
The hardware designers \emph{have} been working on this
problem, and have consulted with no less a luminary than
the late physicist Stephen Hawking.
@@ -423,9 +426,24 @@ thousand clock cycles.
\Cref{sec:cpu:Hardware Free Lunch?}
looks at what else hardware designers might be
able to do to ease the plight of parallel programmers.
-}\QuickQuizEnd
-
-\QuickQuizLabel{\QspeedOfLightAtoms}
+}\QuickQuizEndB
+%
+\QuickQuizE{
+ \Cref{tab:cpu:Performance of Synchronization Mechanisms on 16-CPU 2.8GHz Intel X5550 (Nehalem) System}
+ in the answer to \QuickQuizARef{\QspeedOfLightAtoms} says that
+ In-Core CAS is faster than both of Same-CPU CAS and In-Core Blind CAS\@.
+ What is happening there?
+}\QuickQuizAnswerE{
+ I \emph{was} surprised by the data I obtained and did a rigorous
+ check of their validity.
+ I got the same result persistently.
+ One theory that might explain the observation would be:
+ The two threads in the core are able to overlap their accesses,
+ while the single CPU must do everything sequentially.
+ Unfortunately, there seems to be no public documentation explaining
+ why the Intel X5550 (Nehalem) system behaved like that.
+}\QuickQuizEndE
+} % End of \QuickQuizSeries
\begin{table}
\rowcolors{1}{}{lightgray}
diff --git a/qqz.sty b/qqz.sty
index a3a9f22d1ba9..5c7eb5340194 100644
--- a/qqz.sty
+++ b/qqz.sty
@@ -156,6 +156,11 @@
% To create a macro referencing the previously defined quick quiz:
% \QuickQuizLabel{\QQname}
+%
+% When labeling a QQz inside \QuickQuizSeries{}, use
+% \QuickQuizLabelRel{\QQname}{rel}
+% in front of the series.
+%
% To reference the macro in the text:
% \QuickQuizRef{\QQname}
% To reference the answer of the macro in the text:
@@ -166,6 +171,11 @@
\newcommand{\QuickQuizLabel}[1]{
\edef#1{\thechapter.\thequickquizctrP}
}
+\newcommand{\QuickQuizLabelRel}[2]{
+ \addtocounter{quickquizctrP}{#2}
+ \QuickQuizLabel{#1}
+ \addtocounter{quickquizctrP}{-#2}
+}
\newcommand{\QuickQuizRef}[1]{%
\hyperref[QQ.#1]{Quick Quiz~#1}%
}
@@ -176,6 +186,12 @@
\newcommand{\QuickQuizLabel}[1]{
\edef#1{\thechapter.\thequickquizctr}
}
+\newcommand{\QuickQuizLabelRel}[2]{
+ \addtocounter{quickquizctr}{#2}
+ \QuickQuizLabel{#1}
+ \addtocounter{quickquizctr}{-#2}
+}
+
\newcommand{\QuickQuizRef}[1]{%
\hyperref[QQ.#1]{Quick Quiz~#1}%
}
base-commit: 14440e232cc1b2580dc1a73f873dc29fe3aea02b
--
2.25.1
