Glossary pages now have definitions of "Multi-Copy Atomic",
"Non-Multi-Copy Atomic", and "Other Multi-Copy Atomic".
For consistency, change the latter to "Other-Multi-Copy Atomic".
Index markers added in commit 0685b56dfb58 ("memorder: Add index
markers for "multicopy atomicity"") result in different items
in Index pages:
- Multicopy atomicity
- full
- non-
- other-
Change them so that they land under "Multi-copy atomic".
Also, as "full(y) multi-copy atomic" is synonymous with "multi-copy atomic",
make their index markers plain "Multi-copy atomic".
After this change, Index pages will have just:
- Multi-copy atomic
- non-
- other-
In single-level index pages (with the initial value of boolean
"indexhier" changed to "false" in perfbook-lt.tex), there were
terms with an extra space as listed below:
- Anti- heisenbug
- Non- multicopy atomicity
- Other- multicopy atomicity
To suppress those spaces, Define \IXhy, \IXalthy, and their variants
and use them for index marking. The "hy" comes from "_hy_phen".
Signed-off-by: Akira Yokosawa <[email protected]>
---
debugging/debugging.tex | 3 +--
glossary.tex | 8 ++++----
indexsee.tex | 2 ++
memorder/memorder.tex | 24 +++++++++++++-----------
perfbook-lt.tex | 8 ++++++++
5 files changed, 28 insertions(+), 17 deletions(-)
diff --git a/debugging/debugging.tex b/debugging/debugging.tex
index 0fcf4541..169626bd 100644
--- a/debugging/debugging.tex
+++ b/debugging/debugging.tex
@@ -1701,8 +1701,7 @@ more easily reproduced!
If the field of physics inspired the name of this problem, it is only
fair that the field of physics should inspire the solution.
Fortunately, particle physics is up to the task:
-Why not create an \IXBalth{anti-heisenbug}{anti-}{heisenbug}
-to annihilate the heisenbug?
+Why not create an \IXBhy{anti-}{heisenbug} to annihilate the heisenbug?
Or, perhaps more accurately, to annihilate the heisen-ness of
the heisenbug?
Although producing an anti-heisenbug for a given heisenbug is more an
diff --git a/glossary.tex b/glossary.tex
index 2d295acf..9ded12f8 100644
--- a/glossary.tex
+++ b/glossary.tex
@@ -482,7 +482,7 @@
Note that this means that if a CPU does a write to a given
location followed immediately by a read from that same location,
it might take significant time for that read to complete.
- (See also ``other multi-copy atomic'' and ``non-multi-copy atomic''.)
+ (See also ``other-multi-copy atomic'' and ``non-multi-copy atomic''.)
\item[\IXG{Mutual-Exclusion Mechanism}:]
A software abstraction that regulates threads' access to
``critical sections'' and corresponding data.
@@ -512,11 +512,11 @@
of the stronger forward-progress guarantees, usually wait free or
lock free, but sometimes also obstruction free.
(See ``Non-blocking''.)
-\item[\IXG{Non-Multi-Copy Atomic}:]
+\item[\IXGhy{Non-}{Multi-Copy Atomic}:]
A memory-consistency model in which, in the absence of explicit
memory-ordering instructions, writes to memory may appear to
occur in different orders from the perspective of different CPUs.
- (See also ``multi-copy atomic'' and ``other multi-copy atomic''.)
+ (See also ``multi-copy atomic'' and ``other-multi-copy atomic''.)
\item[NUCA:]\glsuseriii{nuca}
Non-uniform cache architecture, where groups of CPUs share
caches and/or store buffers.
@@ -539,7 +539,7 @@
A forward-progress guarantee in which, in the absence of
contention, every thread makes progress within a finite
period of time.
-\item[\IXG{Other Multi-Copy Atomic}:]
+\item[\IXGhy{Other-}{Multi-Copy Atomic}:]
A memory-consistency model in which a given set of writes to
memory appear to occur in an order consistent with a single
global order, but only from the perspective of CPUs that did
diff --git a/indexsee.tex b/indexsee.tex
index a77e9d64..85527197 100644
--- a/indexsee.tex
+++ b/indexsee.tex
@@ -22,6 +22,8 @@
\index{Memory-barrier latency|see{Latency, memory-barrier}}
\index{Memory-barrier overhead|see{Overhead, memory-barrier}}
\index{Message latency|see{Latency, message}}
+\index{Non-multi-copy atomic|see{Multi-copy atomic, non-}}
+\index{Other-multi-copy atomic|see{Multi-copy atomic, other-}}
\index{Process consistency|see{Consistency, process}}
\index{RCU read-side critical section|see{Critical section, RCU read-side}}
\index{Read memory barrier|see{Memory barrier, read}}
diff --git a/memorder/memorder.tex b/memorder/memorder.tex
index 1a367509..3fab79a5 100644
--- a/memorder/memorder.tex
+++ b/memorder/memorder.tex
@@ -2273,7 +2273,8 @@ accommodate only one store at a time, then any pair of
CPUs would
agree on the order of all stores that they observed.
Unfortunately, building a computer system as shown in the figure, without
store buffers or even caches, would result in glacially slow computation.
-Most CPU vendors interested in providing \IXB{multicopy atomicity} therefore
+Most CPU vendors interested in providing
+\IXBalt{multicopy atomicity}{multi-copy atomic} therefore
instead provide the slightly weaker
\emph{other-multicopy atomicity}~\cite[Section B2.3]{ARMv8A:2017},
which excludes the CPU doing a given store from the requirement that all
@@ -2287,7 +2288,7 @@ CPUs agree on the order of all stores.\footnote{
\cpageref{tab:memorder:Summary of Memory Ordering}.}
This means that if only a subset of CPUs are doing stores, the
other CPUs will agree on the order of stores, hence the ``other''
-in ``\IXBalth{other-multicopy atomicity}{other-}{multicopy atomicity}''.
+in ``\IXBalthy{other-multicopy atomicity}{other-}{multi-copy atomic}''.
Unlike multicopy-atomic platforms, within other-multicopy-atomic platforms,
the CPU doing the store is permitted to observe its
store early, which allows its later loads to obtain the newly stored
@@ -2332,7 +2333,7 @@ value directly from the store buffer, which improves
performance.
Perhaps there will come a day when all platforms provide some flavor
of multi-copy atomicity, but
-in the meantime, \IXalth{non-multicopy-atomic}{non-}{multicopy atomicity}
+in the meantime, \IXalthy{non-multicopy-atomic}{non-}{multi-copy atomic}
platforms do exist, and so software must deal with them.
\begin{listing}
@@ -3448,7 +3449,7 @@ If you wish, you can click on ``Undo'' to explore other
options or
click on ``Reset'' to start over.
It can be very helpful to carry out these steps in different orders
to better understand how a
-\IXalth{non-multicopy-atomic}{non-}{multicopy atomicity}
+\IXalthy{non-multicopy-atomic}{non-}{multi-copy atomic}
architecture operates.
\QuickQuiz{
@@ -5366,13 +5367,14 @@ sequential consistency.
Performance considerations have dictated that no modern mainstream
system is sequentially consistent.
-The next three rows cover \IX{multicopy atomicity}, which was defined in
+The next three rows cover \IXalt{multicopy atomicity}{multi-copy atomic},
+which was defined in
\cref{sec:memorder:Multicopy Atomicity}.
The first is full-up (and rare)
-\IXalth{multicopy atomicity}{full}{multicopy atomicity}, the second is the
-weaker \IXalth{other-multicopy atomicity}{other-}{multicopy atomicity},
+\IXalt{multicopy atomicity}{multi-copy atomic}, the second is the
+weaker \IXalthy{other-multicopy atomicity}{other-}{multi-copy atomic},
and the third is the weakest
-\IXalth{non-multicopy atomicity}{non-}{multicopy atomicity}.
+\IXalthy{non-multicopy atomicity}{non-}{multi-copy atomic}.
The next row, ``Non-Cache Coherent'', covers accesses from multiple
threads to a single variable, which was discussed in
@@ -6075,7 +6077,7 @@ instructions~\cite{PowerPC94,MichaelLyons05a}:
order of these stores.
Not so on PowerPC, even with an \co{lwsync} instruction between each
pair of memory-reference instructions, because PowerPC is
- \IXalth{non-multicopy atomic}{non-}{multicopy atomicity}.
+ \IXalthy{non-multicopy atomic}{non-}{multi-copy atomic}.
\item [\tco{eieio}] (enforce in-order execution of I/O, in case you
were wondering) causes all preceding cacheable stores to appear
to have completed before all subsequent stores.
@@ -6238,7 +6240,7 @@ would indicate.
This exception to the total ordering is needed to allow important
hardware optimizations involving store buffers.
In addition, x86 provides
-\IXalth{other-multicopy atomicity}{other-}{multicopy atomicity}, for example,
+\IXalthy{other-multicopy atomicity}{other-}{multi-copy atomic}, for example,
so that if CPU~0 sees a store by CPU~1, then CPU~0 is guaranteed to see
all stores that CPU~1 saw prior to its store.
Software may use atomic operations to override these hardware optimizations,
@@ -6293,7 +6295,7 @@ It also has strong memory-ordering semantics, as shown in
\cref{tab:memorder:Summary of Memory Ordering}.
In particular, all CPUs will agree on the order of unrelated stores from
different CPUs, that is, the z~Systems CPU family is
-\IXalth{fully multicopy atomic}{full}{multicopy atomicity},
+\IXalt{fully multicopy atomic}{multi-copy atomic},
and is the only commercially available system with this property.
As with most CPUs, the z~Systems architecture does not guarantee a
diff --git a/perfbook-lt.tex b/perfbook-lt.tex
index 580d1f01..fdcb5e8d 100644
--- a/perfbook-lt.tex
+++ b/perfbook-lt.tex
@@ -214,10 +214,12 @@
}
\IfIndexHier{
\newcommand{\indexh}[3]{\ucindexh{#2}{#3}}
+\newcommand{\indexhy}[3]{\ucindexh{#2}{#3}}
\newcommand{\indexhr}[3]{\indexhraw{#2}{#3}}
\newcommand{\indexhmr}[3]{\ucindexhm{#2}{#3}}
}{
\newcommand{\indexh}[3]{\ucindex{#3 #2}}
+\newcommand{\indexhy}[3]{\ucindex{#3#2}}
\newcommand{\indexhr}[3]{\index{#1}}
\newcommand{\indexhmr}[3]{\index{#1}}
}
@@ -235,7 +237,9 @@
\newcommand{\IXhrpl}[2]{\indexhr{#1 #2}{#2}{#1}\hlindex{#1 #2s}}
\newcommand{\IXhmr}[2]{\indexhmr{#1 #2}{#2}{#1}\hlindex{#1 #2}}
\newcommand{\IXhmrpl}[2]{\indexhmr{#1 #2}{#2}{#1}\hlindex{#1 #2s}}
+\newcommand{\IXhy}[2]{\indexhy{#1#2}{#2}{#1}\hlindex{#1#2}}
\newcommand{\IXalth}[3]{\indexh{#1}{#3}{#2}\hlindex{#1}}
+\newcommand{\IXalthy}[3]{\indexhy{#1}{#3}{#2}\hlindex{#1}}
\newcommand{\IXalthr}[3]{\indexhr{#1}{#3}{#2}\hlindex{#1}}
\newcommand{\IXalthmr}[3]{\indexhmr{#1}{#3}{#2}\hlindex{#1}}
% page number in bold face
@@ -253,7 +257,9 @@
\newcommand{\IXBhrpl}[2]{\indexhr{#1 #2|BF}{#2|BF}{#1}\hlindex{#1 #2s}}
\newcommand{\IXBhmr}[2]{\indexhmr{#1 #2|BF}{#2|BF}{#1}\hlindex{#1 #2}}
\newcommand{\IXBhmrpl}[2]{\indexhmr{#1 #2|BF}{#2|BF}{#1}\hlindex{#1 #2s}}
+\newcommand{\IXBhy}[2]{\indexhy{#1#2|BF}{#2|BF}{#1}\hlindex{#1#2}}
\newcommand{\IXBalth}[3]{\indexh{#1|BF}{#3|BF}{#2}\hlindex{#1}}
+\newcommand{\IXBalthy}[3]{\indexhy{#1|BF}{#3|BF}{#2}\hlindex{#1}}
\newcommand{\IXBalthr}[3]{\indexhr{#1|BF}{#3|BF}{#2}\hlindex{#1}}
\newcommand{\IXBalthmr}[3]{\indexhmr{#1|BF}{#3|BF}{#2}\hlindex{#1}}
% page number for Glossary items or the likes
@@ -271,7 +277,9 @@
\newcommand{\IXGhrpl}[2]{\indexhr{#1 #2|GL}{#2|GL}{#1}\hlindex{#1 #2s}}
\newcommand{\IXGhmr}[2]{\indexhmr{#1 #2|GL}{#2|GL}{#1}\hlindex{#1 #2}}
\newcommand{\IXGhmrpl}[2]{\indexhmr{#1 #2|GL}{#2|GL}{#1}\hlindex{#1 #2s}}
+\newcommand{\IXGhy}[2]{\indexhy{#1#2|GL}{#2|GL}{#1}\hlindex{#1#2}}
\newcommand{\IXGalth}[3]{\indexh{#1|GL}{#3|GL}{#2}\hlindex{#1}}
+\newcommand{\IXGalthy}[3]{\indexhy{#1|GL}{#3|GL}{#2}\hlindex{#1}}
\newcommand{\IXGalthr}[3]{\indexhr{#1|GL}{#3|GL}{#2}\hlindex{#1}}
\newcommand{\IXGalthmr}[3]{\indexhmr{#1|GL}{#3|GL}{#2}\hlindex{#1}}
%
--
2.34.1
