fillBytes uses 'memset' to fill a memory area with a required byte value.
I've made all the changes requested by @duncan in https://phabricator.haskell.org/D461:
(all the reviewers from #D461 are preserved)
| Lint Warnings | Excuse: Keeping a style of 'memcpy' and 'memmove' |
| Severity | Location | Code | Message |
|---|---|---|---|
| Warning | libraries/base/Foreign/Marshal/Utils.hs:186 | TXT3 | Line Too Long |
| No Unit Test Coverage |
| Buildable 1836 | |
| Build 1843: GHC Patch Validation (amd64/Linux) |
Hi Everyone,
Sorry to disturb you, do you think the change is reasonable? Do you think you could merge it?
As soon as that one's done, I'll send a patch with @duncan's idea with malloc0, I already have a prototype, implemented it in my FFI tutorial.
@ifesdjeen, as noted in D461#11545, this should undergo discussion on the libraries@ mailing list before we merge it
For reference, here's the proposal thread: http://thread.gmane.org/gmane.comp.lang.haskell.libraries/23347
generally LGTM; got only two doc-related requests:
| libraries/base/Foreign/Marshal/Utils.hs | ||
|---|---|---|
| 173 | Please add an additional -- /Since: 4.8.0.0/ as last-line | |
| 174 | (I'm not sure if this was discussed: is that the most obvious/consistent/natural argument-ordering?) | |
Add 'fillBytes' to Foreign.Marshal.Utils.
fillBytes uses 'memset' to fill a memory area with
a required byte value.
@hvr I've updated changelog.md and added Since into comment block.
As regards the ordering, this one matches the original, and works pretty much the same way as the others (memcpy and memmove). I personally have used it in my project internally for a week or more and it was quite intuitive for me. Do you have objections?
No, no objections. I just wanted to make sure that ordering was chosen deliberately.
| libraries/base/changelog.md | ||
|---|---|---|
| 103 ↗ | (On Diff #1594) | Minor issue: Those aren't backquotes, which means this won't be rendered as monospace font |