| Add Proposal | Add Analysis | Edit Class, Environment, or Release |
| Number | 254
|
| Category | errata
|
| Synopsis | Table 30 should clarify that LHS can be array word or select
|
| State | open
|
| Class | errata-discuss
|
| Arrival-Date | Jan 09 2003
|
| Originator | Shalom.Bresticker@motorola.com
|
| Release | 2001b: 3.10, 4.2, 6, 9.2, 12.3.9.2
|
| Environment |
http://www.boydtechinc.com/btf/archive/btf_2002/1809.html See #319. |
| Description |
This issue comes from the thread starting at http://www.boydtechinc.com/btf/archive/btf_2002/1809.html Table 30 says that LHS of assignments can be scalars, vectors, bit-selects, or part-selects (with some differences between continuous assignments and procedural assignments). It also says that LHS of procedural assignments can be "memory word". This goes back to 1364-1995, when arrays could only be 2-D, and only regs (not nets), and you could only access entire memory word, not bit-select or part-select. Table should clarify that LHS of continuous assignment can be array word (with constant select), or constant bit-select or constant part-select of array word with constant selects. Similarly, LHS of procedural assignments can be array word, or bit-select or part-select of array word, even with non-constant selects. This is relevant also to 2nd paragraph of 9.2, which again talks about a "memory word", a leftover from 1995 text. Also, the description of a "structural net expression", twice in 12.3.9.2 also needs to be updated. All this is connected to issues #75 and 198 as well. Also, the 3rd paragraph in 4.2 also mentions a "memory word". Again, it should be an array word, bit-select, or part-select. Actually, I would like to see sections 3.10 and 4.2.2 clarify better that an array element is a vector/scalar (depending on its declaration) just like any other and can be used anywhere they can, with the same restrictions, unless otherwise stated. Finally, the entire standard should be searched to see whether there are any other archaic references to "memory" or "memories". |
| Fix |
|
| Audit-Trail |
From: "Brad Pierce" <Brad.Pierce@synopsys.com> To: <etf-bugs@boyd.com> Cc: Subject: Re: errata/254: Table 30 should clarify that LHS can be array word/bit-select/part-select Date: Thu, 9 Jan 2003 13:21:21 -0800 The nonterminal "memory_identifier" appears only in A.9.3. There are references to "memories" in the "Memories" section 3.10.3. -- Brad From: Shalom.Bresticker@motorola.com To: Brad Pierce <Brad.Pierce@synopsys.com> Cc: etf-bugs@boyd.com Subject: Re: errata/254: Table 30 should clarify that LHS can be array word/bit-select/part-select Date: Thu, 9 Jan 2003 23:37:04 +0200 (IST) > The nonterminal "memory_identifier" appears only in A.9.3. We already voted to delete that. > There are references to "memories" in the "Memories" section 3.10.3. That is deliberate, due to references in the PLI that we do not want to delete. From: Shalom Bresticker <Shalom.Bresticker@motorola.com> To: etf-bugs@boyd.com Cc: Subject: Re: errata/254: Table 30 should clarify that LHS can be array word/bit-select/part-select Date: Thu, 03 Apr 2003 15:04:44 +0300 #319 largely duplicates 254. 319 should be incorporated into 254. |
| Unformatted |
word/bit-select/part-select |
Hosted by Boyd Technology