Typos in Chapters 1, 2, 3 and 4

[nmnobre]

Hi,

I hope this is the right place to report minor issues, please do let me know otherwise.

• In the section on "Amdahl and Gustafson", last paragraph, second sentence, I believe it should read "Parallelism would not /be/ nearly as important [...]" (p.10)

• In the section on "Scaling", first sentence, it reads "The word "scaling" appeared in our prior discussion.", but this doesn't seem to be the case (p.11)

• In the section on "Device Selectors", after the second table, I believe it should read "queue myQueue{ gpu_selector_v };" (p.47)

• In the code of Figure 2-19, the return type for memset and memcpy are incorrect, should be void, not event (p.57) (addressed in Address minor issues in Figures 2-19 and 3-17 #138)

• In the code of Figure 3-6, there's a mismatch between the actual names of the array and buffer and the references to them in code comments (p.79) (addressed in Fix comments in Figure 3-6 #136)

• In the section on "In-order Queues", third sentence, I believe it should read "Their intuitive task ordering means that in-order queues /have/ an advantage of simplicity [...]" (p.83)

• In the section on "Explicit Dependences with Events", first sentence, it reads "Explicit dependences between tasks look like the examples we have seen (Figure 3-8) [...]" but so do implicit dependences, so I'd personally just remove this sentence (p.84)

• In Figure 3-14, we represent a read on the host as if it were a (CG) task to be able to draw the dependence edge, which I think is fine even if it's a slight abuse of notation; but then we do not represent the host read(s) in Figure 3-16, making the whole thing slightly inconsistent (p.91)

• In the section on "Implicit Dependences with Accessors", last paragraph, I believe it should read "The WAW dependence on buffer b_buf found in kernel /rewriteB/ functions similarly." (p.91)

• In the section on "Handler Class: Key Members", the whole section, but in particular this sentence "This is as good a place as any to lay them out.", just reads like we've forgotten that we've already presented this class in Figure 2-19. There's a reference to Chapter 2, but only for the queue class, whose members were presented as shortcuts of the handler class (p.93)

• In the code of Figure 3-17, the indentation for depends_on and fill is inconsistent with the other members. The first parameter of the range parallel_for is not a valid C++ identifier (p.94) (addressed in Address minor issues in Figures 2-19 and 3-17 #138)

• In Figure 4-19, the get_id()/get_linear_id() members of the group class should be get_group_id()/get_group_linear_id() instead (p.125) (addressed in Fix member function names in Figure 4-19 #139)

Cheers,
-Nuno