[PATCH -perfbook 0/7] Collection of minor fixes and updates

[PATCH -perfbook 0/7] Collection of minor fixes and updates

[Akira Yokosawa]

Hi Paul,

This patch set is a collection of minor fixups I have made
during your traveling.

Patches #1 and #2 are fixups to the \qco{} changes by SJ.

Patch #3 adds a couple of acronym entries.

Patches #4 and #5 follow up code changes made by Alan.

Patch #6 tweaks width params for TOC pages.

Patch #7 fixes a false positive check in precheck.sh WRT the
date command.

        Thanks, Akira

--
Akira Yokosawa (7):
  formal/spinhint: Don't use \qco{} for long message
  formal/ppcmem: Use \qco{} as intended
  Add acronyms of LKMM and KCSAN
  CodeSamples/count: Adjust coding style
  count: Adjust scaling factor of Listing 5.18
  Tweak TOC width params
  precheck.sh: Export LC_TIME

 CodeSamples/count/count_lim.c           |  5 +++--
 CodeSamples/count/count_lim_atomic.c    |  5 +++--
 CodeSamples/count/count_lim_sig.c       | 23 +++++++++++------------
 CodeSamples/count/count_stat_eventual.c |  3 +--
 CodeSamples/count/count_tstat.c         |  3 ++-
 count/count.tex                         |  7 +++----
 debugging/debugging.tex                 |  2 +-
 formal/axiomatic.tex                    |  4 ++--
 formal/ppcmem.tex                       |  2 +-
 formal/spinhint.tex                     |  2 +-
 glsdict.tex                             |  2 ++
 memorder/memorder.tex                   |  9 ++++-----
 perfbook-lt.tex                         | 21 +++++++++++++++++++--
 utilities/precheck.sh                   |  2 +-
 14 files changed, 54 insertions(+), 36 deletions(-)


base-commit: 90a7f29b4961b8d58e2bf1ee6a64c97bb74ee926
-- 
2.25.1

[PATCH -perfbook 1/7] formal/spinhint: Don't use \qco{} for long message

[Akira Yokosawa]

\qco{} is intended to be used for short snippets and suppresses line
breaks in its argument by using \nbco{} in its definition.
Commit 88fd12717c61 ("formal/spinhint: Use \qco{} instead of``\co{}''")
applied \qco{} for a long error message.
It causes "Underfull \hbox" of large badness preceding it.

In perfbook-1c.log and perfbook-eb.log:

    Underfull \hbox (badness 10000) in paragraph at lines 277--281
    []\T1/ntxtlf/m/n/10 (+20) If you see a mes-sage from [][][][] say-ing:
     []

In perfbook.log:

    Underfull \hbox (badness 2495) in paragraph at lines 277--281
    []\T1/ntxtlf/m/n/10 (+20) If you see a mes-sage from [][][][] say-ing:
     []

Revert the change and permit line breaks in the middle of the quoted
message.

Signed-off-by: Akira Yokosawa <akiyks@gmail.com>
Cc: SeongJae Park <sj38.park@gmail.com>
---
 formal/spinhint.tex | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/formal/spinhint.tex b/formal/spinhint.tex
index 9bb5f5749470..16764aef8620 100644
--- a/formal/spinhint.tex
+++ b/formal/spinhint.tex
@@ -275,7 +275,7 @@ Given a source file \path{qrcu.spin}, one can use the following commands:
 	10.5\,GB of memory even with the \co{-DCOLLAPSE} flag.
 
 	If you see a message from \co{./pan} saying:
-	\qco{error: max search depth too small}, you need to increase
+	``\co{error: max search depth too small}'', you need to increase
 	the maximum depth by a \co{-mN} option for a complete search.
 	The default is \co{-m10000}.
 
-- 
2.25.1

Re: [PATCH -perfbook 1/7] formal/spinhint: Don't use \qco{} for long message

[SeongJae Park]

On Sun, 30 Apr 2023 08:05:30 +0900 Akira Yokosawa <akiyks@gmail.com> wrote:

> \qco{} is intended to be used for short snippets and suppresses line
> breaks in its argument by using \nbco{} in its definition.
> Commit 88fd12717c61 ("formal/spinhint: Use \qco{} instead of``\co{}''")
> applied \qco{} for a long error message.
> It causes "Underfull \hbox" of large badness preceding it.
> 
> In perfbook-1c.log and perfbook-eb.log:
> 
>     Underfull \hbox (badness 10000) in paragraph at lines 277--281
>     []\T1/ntxtlf/m/n/10 (+20) If you see a mes-sage from [][][][] say-ing:
>      []
> 
> In perfbook.log:
> 
>     Underfull \hbox (badness 2495) in paragraph at lines 277--281
>     []\T1/ntxtlf/m/n/10 (+20) If you see a mes-sage from [][][][] say-ing:
>      []
> 
> Revert the change and permit line breaks in the middle of the quoted
> message.

Thank you for fixing this, Akira!

> 
> Signed-off-by: Akira Yokosawa <akiyks@gmail.com>
> Cc: SeongJae Park <sj38.park@gmail.com>

Reviewed-by: SeongJae Park <sj38.park@gmail.com>


Thanks,
SJ

> ---
>  formal/spinhint.tex | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/formal/spinhint.tex b/formal/spinhint.tex
> index 9bb5f5749470..16764aef8620 100644
> --- a/formal/spinhint.tex
> +++ b/formal/spinhint.tex
> @@ -275,7 +275,7 @@ Given a source file \path{qrcu.spin}, one can use the following commands:
>  	10.5\,GB of memory even with the \co{-DCOLLAPSE} flag.
>  
>  	If you see a message from \co{./pan} saying:
> -	\qco{error: max search depth too small}, you need to increase
> +	``\co{error: max search depth too small}'', you need to increase
>  	the maximum depth by a \co{-mN} option for a complete search.
>  	The default is \co{-m10000}.
>  
> -- 
> 2.25.1
> 
> 

[PATCH -perfbook 2/7] formal/ppcmem: Use \qco{} as intended

[Akira Yokosawa]

Commit f87fadc4c8ea ("formal/ppcmem: Use \qco{} instead of ``\co{}''")
did not use \qco{} in one of the hunks.

Use \qco{} as intended.

Signed-off-by: Akira Yokosawa <akiyks@gmail.com>
Cc: SeongJae Park <sj38.park@gmail.com>
---
 formal/ppcmem.tex | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/formal/ppcmem.tex b/formal/ppcmem.tex
index 10526bca88ae..9d7704fa43d0 100644
--- a/formal/ppcmem.tex
+++ b/formal/ppcmem.tex
@@ -99,7 +99,7 @@ exists						@lnlbl[assert:b]
 
 \begin{fcvref}[ln:formal:PPCMEM Litmus Test]
 In the example, \clnref{type} identifies the type of system (\qco{ARM} or
-\co{PPC}) and contains the title for the model.
+\qco{PPC}) and contains the title for the model.
 \Clnref{altname} provides a place for an
 alternative name for the test, which you will usually want to leave
 blank as shown in the above example.
-- 
2.25.1

Re: [PATCH -perfbook 2/7] formal/ppcmem: Use \qco{} as intended

[SeongJae Park]

On Sun, 30 Apr 2023 08:07:18 +0900 Akira Yokosawa <akiyks@gmail.com> wrote:

> Commit f87fadc4c8ea ("formal/ppcmem: Use \qco{} instead of ``\co{}''")
> did not use \qco{} in one of the hunks.
> 
> Use \qco{} as intended.

Nice catch!  Thank you for this fix!

> 
> Signed-off-by: Akira Yokosawa <akiyks@gmail.com>
> Cc: SeongJae Park <sj38.park@gmail.com>

Reviewed-by: SeongJae Park <sj38.park@gmail.com>


Thanks,
SJ

> ---
>  formal/ppcmem.tex | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/formal/ppcmem.tex b/formal/ppcmem.tex
> index 10526bca88ae..9d7704fa43d0 100644
> --- a/formal/ppcmem.tex
> +++ b/formal/ppcmem.tex
> @@ -99,7 +99,7 @@ exists						@lnlbl[assert:b]
>  
>  \begin{fcvref}[ln:formal:PPCMEM Litmus Test]
>  In the example, \clnref{type} identifies the type of system (\qco{ARM} or
> -\co{PPC}) and contains the title for the model.
> +\qco{PPC}) and contains the title for the model.
>  \Clnref{altname} provides a place for an
>  alternative name for the test, which you will usually want to leave
>  blank as shown in the above example.
> -- 
> 2.25.1
> 
> 

Re: [PATCH -perfbook 2/7] formal/ppcmem: Use \qco{} as intended

[Paul E. McKenney]

On Sat, Apr 29, 2023 at 04:41:37PM -0700, SeongJae Park wrote:
> On Sun, 30 Apr 2023 08:07:18 +0900 Akira Yokosawa <akiyks@gmail.com> wrote:
> 
> > Commit f87fadc4c8ea ("formal/ppcmem: Use \qco{} instead of ``\co{}''")
> > did not use \qco{} in one of the hunks.
> > 
> > Use \qco{} as intended.
> 
> Nice catch!  Thank you for this fix!
> 
> > 
> > Signed-off-by: Akira Yokosawa <akiyks@gmail.com>
> > Cc: SeongJae Park <sj38.park@gmail.com>
> 
> Reviewed-by: SeongJae Park <sj38.park@gmail.com>

I pulled in these along with SeongJae's patches, thank you both!

							Thanx, Paul

> Thanks,
> SJ
> 
> > ---
> >  formal/ppcmem.tex | 2 +-
> >  1 file changed, 1 insertion(+), 1 deletion(-)
> > 
> > diff --git a/formal/ppcmem.tex b/formal/ppcmem.tex
> > index 10526bca88ae..9d7704fa43d0 100644
> > --- a/formal/ppcmem.tex
> > +++ b/formal/ppcmem.tex
> > @@ -99,7 +99,7 @@ exists						@lnlbl[assert:b]
> >  
> >  \begin{fcvref}[ln:formal:PPCMEM Litmus Test]
> >  In the example, \clnref{type} identifies the type of system (\qco{ARM} or
> > -\co{PPC}) and contains the title for the model.
> > +\qco{PPC}) and contains the title for the model.
> >  \Clnref{altname} provides a place for an
> >  alternative name for the test, which you will usually want to leave
> >  blank as shown in the above example.
> > -- 
> > 2.25.1
> > 
> > 

[PATCH -perfbook 3/7] Add acronyms of LKMM and KCSAN

[Akira Yokosawa]

LKMM and KCSAN deserve entries in the list of Acronyms
Add them.

Signed-off-by: Akira Yokosawa <akiyks@gmail.com>
---
 debugging/debugging.tex | 2 +-
 formal/axiomatic.tex    | 4 ++--
 glsdict.tex             | 2 ++
 memorder/memorder.tex   | 9 ++++-----
 4 files changed, 9 insertions(+), 8 deletions(-)

diff --git a/debugging/debugging.tex b/debugging/debugging.tex
index 191967baa810..42d69a9337fa 100644
--- a/debugging/debugging.tex
+++ b/debugging/debugging.tex
@@ -764,7 +764,7 @@ helpful in production.
 
 An especially bad parallel-code something is unexpected concurrent
 access to data.
-The Kernel Concurrency Sanitizer (KCSAN)~\cite{JonathanCorbet2019KCSAN}
+The \IXBacrfst{kcsan}~\cite{JonathanCorbet2019KCSAN}
 uses existing markings such as \co{READ_ONCE()} and \co{WRITE_ONCE()}
 to determine which concurrent accesses deserve warning messages.
 KCSAN has a significant false-positive rate, especially from the
diff --git a/formal/axiomatic.tex b/formal/axiomatic.tex
index c970d3bb8382..04adec21c010 100644
--- a/formal/axiomatic.tex
+++ b/formal/axiomatic.tex
@@ -142,8 +142,8 @@ Axiomatic approaches may also be applied to higher-level
 languages and also to higher-level synchronization primitives, as
 exemplified by the lock-based litmus test shown in
 \cref{lst:formal:Locking Example} (\path{C-Lock1.litmus}).
-This litmus test can be modeled by the Linux kernel memory model
-(LKMM)~\cite{Alglave:2018:FSC:3173162.3177156,LucMaranget2018lock.cat}.
+This litmus test can be modeled by
+the \IXacrf{lkmm}~\cite{Alglave:2018:FSC:3173162.3177156,LucMaranget2018lock.cat}.
 As expected, the \co{herd} tool's output features the string \co{Never},
 correctly indicating that \co{P1()} cannot see \co{x} having a value
 of one.\footnote{
diff --git a/glsdict.tex b/glsdict.tex
index d14688590bbb..ce48154acf5f 100644
--- a/glsdict.tex
+++ b/glsdict.tex
@@ -230,6 +230,8 @@
 \newacronym{gpgpu}{GPGPU}{general\-/purpose graphical processing unit}
 \newacronym{irq}{IRQ}{interrupt request}
 \newacronym{ipi}{IPI}{inter\-/processor interrupt}
+\newacronym{kcsan}{KCSAN}{kernel concurrency sanitizer}
+\newacronym{lkmm}{LKMM}{Linux kernel memory consistency model}
 \newacronym{mpi}{MPI}{Message Passing Interface}
 \newacronym{nbs}{NBS}{non-blocking synchronization}
 \newacronym{nmi}{NMI}{non-maskable interrupt}
diff --git a/memorder/memorder.tex b/memorder/memorder.tex
index f116ab81aceb..47bde79b5abd 100644
--- a/memorder/memorder.tex
+++ b/memorder/memorder.tex
@@ -60,11 +60,10 @@ provides some reliable intuitions and useful rules of thumb.
 	along with a paper describing differences between the C11 and
 	Linux memory models~\cite{PaulEMcKenney2016P0124R6-LKMM}.
 
-	The kernel concurrency sanitizer
-	(KCSAN)~\cite{MarcoElver2020FearDataRaceDetector1,MarcoElver2020FearDataRaceDetector2},
+	The \IXacrf{kcsan}~\cite{MarcoElver2020FearDataRaceDetector1,MarcoElver2020FearDataRaceDetector2},
 	based in part on
 	RacerD~\cite{SamBlackshear2018RacerD}
-	and implementing LKMM, has also been added to the Linux kernel
+	and implementing \IXacr{lkmm}, has also been added to the Linux kernel
 	and is now heavily used.
 
 	Finally, there are now better ways of describing LKMM.
@@ -3749,7 +3748,7 @@ For their part, weakly ordered systems might well choose to execute
 the memory-barrier instructions required to guarantee both orderings,
 possibly simplifying code making advanced use of combinations of locked
 and lockless accesses.
-However, as noted earlier, LKMM chooses not to provide these additional
+However, as noted earlier, \IXacr{lkmm} chooses not to provide these additional
 orderings, in part to avoid imposing performance penalties on the simpler
 and more prevalent locking use cases.
 Instead, the \co{smp_mb__after_spinlock()} and \co{smp_mb__after_unlock_lock()}
@@ -4997,7 +4996,7 @@ None of these instructions exactly match the semantics of Linux's
 The \co{DMB} and \co{DSB} instructions have a recursive definition
 of accesses ordered before and after the barrier, which has an effect
 similar to that of \Power{}'s cumulativity, both of which are
-stronger than LKMM's cumulativity described in
+stronger than \IXacr{lkmm}'s cumulativity described in
 \cref{sec:memorder:Cumulativity}.
 
 \ARM\ also implements control dependencies, so that if a conditional
-- 
2.25.1

[PATCH -perfbook 4/7] CodeSamples/count: Adjust coding style

[Akira Yokosawa]

Recent removal of smp_mb() in count_lim.c made a couple of if blocks
one liner.  Remove unneeded braces as per the Linux coding style.

While at it, make similar style fixes under CodeSamples/count as well
as align line-label tail comments.

Signed-off-by: Akira Yokosawa <akiyks@gmail.com>
Cc: Alan Huang <mmpgouride@gmail.com>
---
 CodeSamples/count/count_lim.c           |  5 +++--
 CodeSamples/count/count_lim_atomic.c    |  5 +++--
 CodeSamples/count/count_lim_sig.c       | 23 +++++++++++------------
 CodeSamples/count/count_stat_eventual.c |  3 +--
 CodeSamples/count/count_tstat.c         |  3 ++-
 count/count.tex                         |  5 ++---
 6 files changed, 22 insertions(+), 22 deletions(-)

diff --git a/CodeSamples/count/count_lim.c b/CodeSamples/count/count_lim.c
index db78445f2395..7e379d582342 100644
--- a/CodeSamples/count/count_lim.c
+++ b/CodeSamples/count/count_lim.c
@@ -79,9 +79,10 @@ static __inline__ unsigned long read_count(void)	//\lnlbl{read:b}
 
 	spin_lock(&gblcnt_mutex);			//\lnlbl{read:acquire}
 	sum = globalcount;				//\lnlbl{read:initsum}
-	for_each_thread(t)				//\lnlbl{read:loop:b}
+	for_each_thread(t) {				//\lnlbl{read:loop:b}
 		if (counterp[t] != NULL)
-			sum += READ_ONCE(*counterp[t]);	//\lnlbl{read:loop:e}
+			sum += READ_ONCE(*counterp[t]);
+	}						//\lnlbl{read:loop:e}
 	spin_unlock(&gblcnt_mutex);			//\lnlbl{read:release}
 	return sum;					//\lnlbl{read:return}
 }							//\lnlbl{read:e}
diff --git a/CodeSamples/count/count_lim_atomic.c b/CodeSamples/count/count_lim_atomic.c
index 20d9612d6992..8c4dcc75989e 100644
--- a/CodeSamples/count/count_lim_atomic.c
+++ b/CodeSamples/count/count_lim_atomic.c
@@ -171,11 +171,12 @@ unsigned long read_count(void)				//\lnlbl{b}
 
 	spin_lock(&gblcnt_mutex);			//\lnlbl{acquire}
 	sum = globalcount;				//\lnlbl{initsum}
-	for_each_thread(t)				//\lnlbl{loop:b}
+	for_each_thread(t) {				//\lnlbl{loop:b}
 		if (counterp[t] != NULL) {
 			split_counterandmax(counterp[t], &old, &c, &cm);//\lnlbl{split}
 			sum += c;
-		}					//\lnlbl{loop:e}
+		}
+	}						//\lnlbl{loop:e}
 	spin_unlock(&gblcnt_mutex);			//\lnlbl{release}
 	return sum;					//\lnlbl{return}
 }							//\lnlbl{e}
diff --git a/CodeSamples/count/count_lim_sig.c b/CodeSamples/count/count_lim_sig.c
index c2f611976466..8d6c3ca3def5 100644
--- a/CodeSamples/count/count_lim_sig.c
+++ b/CodeSamples/count/count_lim_sig.c
@@ -57,9 +57,8 @@ static void flush_local_count_sig(int unused)	//\lnlbl{flush_sig:b}
 	if (READ_ONCE(theft) != THEFT_REQ)	//\lnlbl{flush_sig:check:REQ}
 		return;				//\lnlbl{flush_sig:return:n}
 	WRITE_ONCE(theft, THEFT_ACK);		//\lnlbl{flush_sig:set:ACK}
-	if (!counting) {			//\lnlbl{flush_sig:check:fast}
+	if (!counting)				//\lnlbl{flush_sig:check:fast}
 		smp_store_release(&theft, THEFT_READY);	//\lnlbl{flush_sig:set:READY}
-	}
 }						//\lnlbl{flush_sig:e}
 
 static void flush_local_count(void)			//\lnlbl{flush:b}
@@ -67,7 +66,7 @@ static void flush_local_count(void)			//\lnlbl{flush:b}
 	int t;
 	thread_id_t tid;
 
-	for_each_tid(t, tid)				//\lnlbl{flush:loop:b}
+	for_each_tid(t, tid) {				//\lnlbl{flush:loop:b}
 		if (theftp[t] != NULL) {		//\lnlbl{flush:skip}
 			if (*countermaxp[t] == 0) {	//\lnlbl{flush:checkmax}
 				WRITE_ONCE(*theftp[t], THEFT_READY);//\lnlbl{flush:READY}
@@ -75,7 +74,8 @@ static void flush_local_count(void)			//\lnlbl{flush:b}
 			}
 			WRITE_ONCE(*theftp[t], THEFT_REQ);//\lnlbl{flush:REQ}
 			pthread_kill(tid, SIGUSR1);	//\lnlbl{flush:signal}
-		}					//\lnlbl{flush:loop:e}
+		}
+	}						//\lnlbl{flush:loop:e}
 	for_each_tid(t, tid) {				//\lnlbl{flush:loop2:b}
 		if (theftp[t] == NULL)			//\lnlbl{flush:skip:nonexist}
 			continue;			//\lnlbl{flush:next2}
@@ -88,7 +88,7 @@ static void flush_local_count(void)			//\lnlbl{flush:b}
 		*counterp[t] = 0;
 		globalreserve -= *countermaxp[t];
 		*countermaxp[t] = 0;			//\lnlbl{flush:thiev:e}
-		smp_store_release(theftp[t], THEFT_IDLE);	//\lnlbl{flush:IDLE}
+		smp_store_release(theftp[t], THEFT_IDLE);//\lnlbl{flush:IDLE}
 	}						//\lnlbl{flush:loop2:e}
 }							//\lnlbl{flush:e}
 
@@ -114,7 +114,7 @@ int add_count(unsigned long delta)			//\lnlbl{b}
 
 	WRITE_ONCE(counting, 1);			//\lnlbl{fast:b}
 	barrier();					//\lnlbl{barrier:1}
-	if (smp_load_acquire(&theft) <= THEFT_REQ &&		//\lnlbl{check:b}
+	if (smp_load_acquire(&theft) <= THEFT_REQ &&	//\lnlbl{check:b}
 	    countermax - counter >= delta) {		//\lnlbl{check:e}
 		WRITE_ONCE(counter, counter + delta);	//\lnlbl{add:f}
 		fastpath = 1;				//\lnlbl{fasttaken}
@@ -122,9 +122,8 @@ int add_count(unsigned long delta)			//\lnlbl{b}
 	barrier();					//\lnlbl{barrier:2}
 	WRITE_ONCE(counting, 0);			//\lnlbl{clearcnt}
 	barrier();					//\lnlbl{barrier:3}
-	if (READ_ONCE(theft) == THEFT_ACK) {		//\lnlbl{check:ACK}
-		smp_store_release(&theft, THEFT_READY);		//\lnlbl{READY}
-	}
+	if (READ_ONCE(theft) == THEFT_ACK)		//\lnlbl{check:ACK}
+		smp_store_release(&theft, THEFT_READY);	//\lnlbl{READY}
 	if (fastpath)
 		return 1;				//\lnlbl{return:fs}
 	spin_lock(&gblcnt_mutex);			//\lnlbl{acquire}
@@ -160,9 +159,8 @@ int sub_count(unsigned long delta)
 	barrier();
 	WRITE_ONCE(counting, 0);
 	barrier();
-	if (READ_ONCE(theft) == THEFT_ACK) {
+	if (READ_ONCE(theft) == THEFT_ACK)
 		smp_store_release(&theft, THEFT_READY);
-	}
 	if (fastpath)
 		return 1;
 	spin_lock(&gblcnt_mutex);
@@ -189,9 +187,10 @@ unsigned long read_count(void)
 
 	spin_lock(&gblcnt_mutex);
 	sum = globalcount;
-	for_each_thread(t)
+	for_each_thread(t) {
 		if (counterp[t] != NULL)
 			sum += READ_ONCE(*counterp[t]);
+	}
 	spin_unlock(&gblcnt_mutex);
 	return sum;
 }
diff --git a/CodeSamples/count/count_stat_eventual.c b/CodeSamples/count/count_stat_eventual.c
index 7157ee0ed9fd..ed081e9102cd 100644
--- a/CodeSamples/count/count_stat_eventual.c
+++ b/CodeSamples/count/count_stat_eventual.c
@@ -50,9 +50,8 @@ void *eventual(void *arg)				//\lnlbl{eventual:b}
 			sum += READ_ONCE(per_thread(counter, t));
 		WRITE_ONCE(global_count, sum);
 		poll(NULL, 0, 1);
-		if (READ_ONCE(stopflag)) {
+		if (READ_ONCE(stopflag))
 			smp_store_release(&stopflag, stopflag + 1);
-		}
 	}
 	return NULL;
 }							//\lnlbl{eventual:e}
diff --git a/CodeSamples/count/count_tstat.c b/CodeSamples/count/count_tstat.c
index 127946f32196..d7b4e335f07b 100644
--- a/CodeSamples/count/count_tstat.c
+++ b/CodeSamples/count/count_tstat.c
@@ -40,9 +40,10 @@ static __inline__ unsigned long read_count(void)
 	int t;
 	unsigned long sum = 0;
 
-	for_each_thread(t)
+	for_each_thread(t) {
 		if (READ_ONCE(counterp[t]) != NULL)
 			sum += READ_ONCE(*counterp[t]);
+	}
 	return sum;
 }
 
diff --git a/count/count.tex b/count/count.tex
index 7b2fde109490..3b8b1c113486 100644
--- a/count/count.tex
+++ b/count/count.tex
@@ -2763,9 +2763,8 @@ write_lock(&mylock);		\lnlbl[acq]
 removing = 1;			\lnlbl[note]
 sub_count(mybias);
 write_unlock(&mylock);		\lnlbl[rel]
-while (read_count() != 0) {	\lnlbl[loop:b]
-	poll(NULL, 0, 1);
-}				\lnlbl[loop:e]
+while (read_count() != 0)	\lnlbl[loop:b]
+	poll(NULL, 0, 1);	\lnlbl[loop:e]
 remove_device();		\lnlbl[remove]
 \end{VerbatimN}
 \end{fcvlabel}
-- 
2.25.1

[PATCH -perfbook 5/7] count: Adjust scaling factor of Listing 5.18

[Akira Yokosawa]

Recent code updates reduce the line count of this snippet.
Use a scaling factor closer to 1.0 for E-Book size builds.

Signed-off-by: Akira Yokosawa <akiyks@gmail.com>
Cc: Alan Huang <mmpgouride@gmail.com>
---
 count/count.tex | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/count/count.tex b/count/count.tex
index 3b8b1c113486..48e3f03d4ce2 100644
--- a/count/count.tex
+++ b/count/count.tex
@@ -2434,7 +2434,7 @@ the fastpath happens before this change of \co{theft} to READY\@.
 \end{fcvref}
 
 \begin{listing}
-\ebresizeverb{.81}{\input{CodeSamples/count/count_lim_sig@migration.fcv}}
+\ebresizeverb{.839}{\input{CodeSamples/count/count_lim_sig@migration.fcv}}
 \caption{Signal-Theft Limit Counter Value-Migration Functions}
 \label{lst:count:Signal-Theft Limit Counter Value-Migration Functions}
 \end{listing}
-- 
2.25.1

[PATCH -perfbook 6/7] Tweak TOC width params

[Akira Yokosawa]

We now have Section 15.5.10, (five digits + two periods).
Default of the book class can not afford that much digits in
subsection count.
Increase width parameters to accomodate it.

Also, as for -sf builds, the Helvetica clone font needs slightly
wider spaces in the "em" unit.  Add settings for -sf builds.

Mention default widths in the comment block for future reference.

Signed-off-by: Akira Yokosawa <akiyks@gmail.com>
---
 perfbook-lt.tex | 21 +++++++++++++++++++--
 1 file changed, 19 insertions(+), 2 deletions(-)

diff --git a/perfbook-lt.tex b/perfbook-lt.tex
index cdac9d495a3a..af5e0bf30032 100644
--- a/perfbook-lt.tex
+++ b/perfbook-lt.tex
@@ -158,11 +158,28 @@
 \setboolean{indexhier}{true}
 \newcommand{\IfIndexHier}[2]{\ifthenelse{\boolean{indexhier}}{#1}{#2}}
 
-% Widen pnumwidth in TOC for ebook size build (more than 1000 pages)
+% Tweak width params of TOC
 \makeatletter
-\IfEbookSize{
+\IfEbookSize{ % for ebook size build (more than 1000 pages)
 \renewcommand*\@pnumwidth{2.2em}
 }{}
+% default params defined in book.sty:
+%  width of chapter (two digits):			1.5em
+%  indent of section:					1.5em
+%  width of section (three digits + one periods):	2.3em
+%  indent of subsection:	  			3.8em
+%  width of subsection (four digits + two periods):	3.2em
+\IfSansSerif{	% sans serif (Helvetica clone)
+		%  to cover section "E.10" and subsection "15.5.10",
+		%  width of section:      2.4em
+		%  width of subsection:   3.7em
+\renewcommand*\l@section{\@dottedtocline{1}{1.5em}{2.4em}}
+\renewcommand*\l@subsection{\@dottedtocline{2}{3.9em}{3.7em}}
+}{		% serif (Times Roman clone)
+		%  to cover subsection "15.5.10",
+		%  width of subsection:   3.4em
+\renewcommand*\l@subsection{\@dottedtocline{2}{3.8em}{3.4em}}
+}
 \makeatother
 
 \IfEbookSize{
-- 
2.25.1

Re: [PATCH -perfbook 6/7] Tweak TOC width params

[Paul E. McKenney]

On Sun, Apr 30, 2023 at 08:22:29AM +0900, Akira Yokosawa wrote:
> We now have Section 15.5.10, (five digits + two periods).
> Default of the book class can not afford that much digits in
> subsection count.
> Increase width parameters to accomodate it.
> 
> Also, as for -sf builds, the Helvetica clone font needs slightly
> wider spaces in the "em" unit.  Add settings for -sf builds.
> 
> Mention default widths in the comment block for future reference.
> 
> Signed-off-by: Akira Yokosawa <akiyks@gmail.com>

And I also pulled in the rest of these patches, thank you!

Huh.  Maybe this is a hint that I should drop one of the CPU families.
Alpha and Itanium would be good candidates except for their historical
significance.  ARM7 and PowerPC are significant in being non-multicopy
atomic.  z Systems is significant in being fully multicopy atomic (as
opposed to merely other-multicopy atomic).  Armv8 and x86 are heavily
used.  MIPS appears to feed directly into LoongArch, which may prove to
be a candidate for addition, as might RISC-V.  That leaves SPARC.

But that means I would be removing one and adding two, so this patch
would still be required.

So it goes!  ;-)

							Thanx, Paul

> ---
>  perfbook-lt.tex | 21 +++++++++++++++++++--
>  1 file changed, 19 insertions(+), 2 deletions(-)
> 
> diff --git a/perfbook-lt.tex b/perfbook-lt.tex
> index cdac9d495a3a..af5e0bf30032 100644
> --- a/perfbook-lt.tex
> +++ b/perfbook-lt.tex
> @@ -158,11 +158,28 @@
>  \setboolean{indexhier}{true}
>  \newcommand{\IfIndexHier}[2]{\ifthenelse{\boolean{indexhier}}{#1}{#2}}
>  
> -% Widen pnumwidth in TOC for ebook size build (more than 1000 pages)
> +% Tweak width params of TOC
>  \makeatletter
> -\IfEbookSize{
> +\IfEbookSize{ % for ebook size build (more than 1000 pages)
>  \renewcommand*\@pnumwidth{2.2em}
>  }{}
> +% default params defined in book.sty:
> +%  width of chapter (two digits):			1.5em
> +%  indent of section:					1.5em
> +%  width of section (three digits + one periods):	2.3em
> +%  indent of subsection:	  			3.8em
> +%  width of subsection (four digits + two periods):	3.2em
> +\IfSansSerif{	% sans serif (Helvetica clone)
> +		%  to cover section "E.10" and subsection "15.5.10",
> +		%  width of section:      2.4em
> +		%  width of subsection:   3.7em
> +\renewcommand*\l@section{\@dottedtocline{1}{1.5em}{2.4em}}
> +\renewcommand*\l@subsection{\@dottedtocline{2}{3.9em}{3.7em}}
> +}{		% serif (Times Roman clone)
> +		%  to cover subsection "15.5.10",
> +		%  width of subsection:   3.4em
> +\renewcommand*\l@subsection{\@dottedtocline{2}{3.8em}{3.4em}}
> +}
>  \makeatother
>  
>  \IfEbookSize{
> -- 
> 2.25.1
> 
> 

[PATCH -perfbook 7/7] precheck.sh: Export LC_TIME

[Akira Yokosawa]

The check script doesn't work properly under non-English locale
settings due to missing "export".

Fix it.

Fixes: 0c852f175f8e ("Add precheck.sh")
Signed-off-by: Akira Yokosawa <akiyks@gmail.com>
---
 utilities/precheck.sh | 2 +-
 1 file changed, 1 insertion(+), 1 deletion(-)

diff --git a/utilities/precheck.sh b/utilities/precheck.sh
index 7de0e9fba328..d98bb0de90ce 100755
--- a/utilities/precheck.sh
+++ b/utilities/precheck.sh
@@ -6,7 +6,7 @@
 #
 # Copyright (C) Akira Yokosawa, 2023
 
-LC_TIME=C
+export LC_TIME=C
 
 : ${SED:=sed}
 : ${DATE:=date}
-- 
2.25.1