On 2016/07/17 08:40:54 +0900, SeongJae Park wrote:
On Sun, 17 Jul 2016, Akira Yokosawa wrote:
Hi,
On 2016/07/16 16:58:49 +0900, SeongJae Park wrote:
Signed-off-by: SeongJae Park <sj38.park@xxxxxxxxx>
---
defer/whichtochoose.tex | 2 +-
1 file changed, 1 insertion(+), 1 deletion(-)
diff --git a/defer/whichtochoose.tex b/defer/whichtochoose.tex
index 648cdc8..884de9d 100644
--- a/defer/whichtochoose.tex
+++ b/defer/whichtochoose.tex
@@ -135,7 +135,7 @@ capable of unconditionally acquiring references.
The entry for sequence locks is ``N/A'' because, again, sequence locks
detect updates rather than acquiring references.
Reference counting and hazard pointers require that traversals be
-restarted from the beginning should a given acquisition fail because
+restarted from the beginning if a given acquisition fail because
I can't get the point of this change. Original sentence seems quite natural
for me.
Well, I thought it would be easier to understand the meaning for poor
English speakers like me, though judging is Paul's right. So, please feel
free to discard this patch if you think it's unnecessary, Paul. BTW, I
think I found one trivial grammar mistake here:
`s/a given acquisition fail/a given acquisition fails/`.
the currrent object might be removed and all of its
possible successsors might be not only be removed, but also freed.
For example, when using reference counting or hazard pointers to
However, there are typos here. We need to modify as follows.
-the currrent object might be removed and all of its
-possible successsors might be not only be removed, but also freed.
+the current object might be removed and all of its
+possible successors might be not only be removed, but also freed.
As I reread the sentence, there is an unnecessary "be" here.
-the currrent object might be removed and all of its
-possible successsors might be not only be removed, but also freed.
+the current object might be removed and all of its
+possible successors might be not only removed, but also freed.
Or we can say:
+the current object might be removed and all of its
+possible successors might not only be removed, but also be freed.
I agree with you that this sentence is fairy complicated.
There might be a better way to rephrase, for example, denoting
the end of that-clause by a comma as appended:
Thanks, Akira
-------------------------------------------------------------
commit 137f7250edbd01b1eab4cbd02449228a949ac2cf
Author: Akira Yokosawa <akiyks@xxxxxxxxx>
Date: Sun Jul 17 10:42:56 2016 +0900
defer/whichtochoose: Rephrase a sentence
This commit adds a comma to denote the end of that-clause.
It also fixes a few typos including the entry of sequence locks
in Table 9.6.
Reported-by: SeongJae Park <sj38.park@xxxxxxxxx>
Signed-off-by: Akira Yokosawa <akiyks@xxxxxxxxx>
diff --git a/defer/whichtochoose.tex b/defer/whichtochoose.tex
index 648cdc8..923c3b3 100644
--- a/defer/whichtochoose.tex
+++ b/defer/whichtochoose.tex
@@ -57,7 +57,7 @@
Reader Reference Acquisition
& Can fail (conditional)
& Can fail (conditional)
- & Unsafe
+ & N/A
& Cannot fail (unconditional) \\
\hline
Memory Footprint
@@ -135,9 +135,9 @@ capable of unconditionally acquiring references.
The entry for sequence locks is ``N/A'' because, again, sequence locks
detect updates rather than acquiring references.
Reference counting and hazard pointers require that traversals be
-restarted from the beginning should a given acquisition fail because
-the currrent object might be removed and all of its
-possible successsors might be not only be removed, but also freed.
+restarted from the beginning should a given acquisition fail, because
+the current object might be removed and all of its
+possible successors might be not only removed, but also freed.
For example, when using reference counting or hazard pointers to
protect a tree traversal, any reference-acquisition failure requires
that the traversal be restarted from the root of the tree.
--
To unsubscribe from this list: send the line "unsubscribe perfbook" in
the body of a message to majordomo@xxxxxxxxxxxxxxx
More majordomo info at http://vger.kernel.org/majordomo-info.html
