Hi Paul,
On Thu, Feb 13, 2020 at 11:19:19AM -0800, Paul E. McKenney wrote:
>On Fri, Feb 14, 2020 at 12:28:24AM +0900, Akira Yokosawa wrote:
>> Hi, Junchang,
>>
>> On Wed, 12 Feb 2020 22:35:15 +0800, Junchang Wang wrote:
>> > Signed-off-by: Junchang Wang <junchangwang@xxxxxxxxx>
>> > ---
>> >
>> > Hi Paul,
>> >
>> > I'm reading the latest version of the perfbook and this patch contains fixes to
>> > some typos in Section Data Structures. Hope it helps.
>> >
>> > Thanks,
>> > --Junchang
>> >
>> > --
>> > datastruct/datastruct.tex | 8 ++++----
>> > 1 file changed, 4 insertions(+), 4 deletions(-)
>> >
>> > diff --git a/datastruct/datastruct.tex b/datastruct/datastruct.tex
>> > index 4b62471..e21952b 100644
>> > --- a/datastruct/datastruct.tex
>> > +++ b/datastruct/datastruct.tex
>> > @@ -700,7 +700,7 @@ Of course, it is quite possible that the differences in lookup performance
>> > are affected by the differences in update rates.
>> > One way to check this is to artificially throttle the update rates of
>> > per-bucket locking and hazard pointers to match that of RCU.
>> > -Doing so does not significantly improve the lookup performace of
>> > +Doing so does not significantly improve the lookup performance of
>> > per-bucket locking, nor does it close the gap between hazard pointers
>> > and RCU.
>> > However, removing the read-side memory barriers from hazard pointers
>> > @@ -800,8 +800,8 @@ scalability, external consistency, or all of the above.
>> > \section{Non-Partitionable Data Structures}
>> > \label{sec:datastruct:Non-Partitionable Data Structures}
>> > %
>> > -\epigraph{Undertake somthing difficult, otherwise you will never grow.}
>> > - {\emph{Ronald E.~Osburn}}
>> > +\epigraph{Undertake something difficult, otherwise you will never grow.}
>> > + {\emph{Ronald E.~Osborn}}
>> >
>> > \begin{figure}[tb]
>> > \centering
>> > @@ -1063,7 +1063,7 @@ Otherwise, a concurrent resize operation has already distributed this
>> > bucket, so line~\lnref{new_hashtbl} proceeds to the new hash table,
>> > line~\lnref{get_newbkt} selects the bucket corresponding to the key,
>> > and line~\lnref{acq_newbkt} acquires the bucket's lock.
>> > -\Clnref{lsp1b}{lsp1e} store the bucket pointer and
>> > +\Clnrefrange{lsp1b}{lsp1e} store the bucket pointer and
>>
>> Thanks for catching this!
>
>And from me as well! Queued and pushed, thank you!
>
Thank you for writing this great book. I'm very happy that I can help
:-)
>> As far as I see, there is no other error in the use of \clnrefrange/\crefrange
>> at the moment.
>> To prevent similar typos in the future, I'll add a check of this pattern somewhere
>> in the build script.
>
>Looking forward to it!
>
>> PS:
>>
>> The email address in the From: field does not match the SOB tag.
>> Is this intentional?
>
>I made the Signed-off-by: field match the From: field. This can be
>changed if need be.
>
I'll use the new email address (junchang2020@xxxxxxxxx) in my
Signed-off-by message in the future. So it's not necessary for you to
change what you have done. Sorry for confusing you and Akira.
Thanks,
--Junchang
> Thanx, Paul
>
>> > pointer-set index into their respective fields in the
>> > \co{ht_lock_state} structure, which again communicates this information to
>> > \co{hashtab_add()}, \co{hashtab_del()}, and \co{hashtab_unlock_mod()}.
>> >
