| Anonymous | Login | 2010-09-02 20:27 PDT |
| Main | My View | View Issues | Docs | Wiki |
| Viewing Issue Simple Details [ Jump to Notes ] [ Wiki ] | [ View Advanced ] [ Issue History ] [ Print ] | ||||||
| ID | Category | Severity | Date Submitted | Last Update | |||
| 0001468 | [SystemVerilog P1800] SV-BC | major | 2006-05-11 16:28 | 2008-06-22 10:48 | |||
| Reporter | Brad Pierce | View Status | public | ||||
| Assigned To | Thomas R Alsop | ||||||
| Priority | immediate | Resolution | fixed | ||||
| Status | closed | Product Version | |||||
| Summary | 0001468: always_latch has same restrictions as always_comb (11.3) | ||||||
| Description |
Doug asks in http://www.eda-stds.org/sv-bc/hm/1878.html [^] whether all of the restrictions of always_comb also apply to always_latch. I think they do, but this is not clear from the text of 11.3. For example, it needs to be clearly stated that variables written on the LHS of assignments in always_latch cannot be written by any other process. For example, it needs to be clearly stated that only the same subset of statement kinds is allowed in always_latch as in always_comb. |
||||||
| Additional Information | |||||||
| Tags | No tags attached. | ||||||
| Type | Errata | ||||||
| Attached Files |
|
||||||
|
|
|||||||
Relationships |
|||||||||||||||||||||
|
|||||||||||||||||||||
Notes |
|
|
shalom (manager) 2006-08-05 19:50 |
The statemeent in 11.2 that, "Statements in an always_comb shall not include those that block", could be misinterpreted to forbid blocking assignments. |
|
Brad Pierce (developer) 2006-08-05 23:59 |
The phrase "blocking statement" is used several times in the LRM without a definition. The language in this section about "those that block" was proposed in http://www.eda-stds.org/sv-bc/hm/1345.html [^] |
|
shalom (manager) 2006-08-06 00:48 |
In 1364, the term "blocking" is used only to refer to blocking assignments. In 1800, it is used in additional contexts. Looks like this should be an additional Mantis, i.e., "What does blocking mean?" |
|
shalom (manager) 2006-11-15 22:40 |
See 225, 1561 |
|
Thomas R Alsop (administrator) 2007-09-17 11:49 |
I have updated the proposal to more clearly word in the always_latch clause that all statements in the 9.2.2.2. (always_comb) apply to it as well and then I emphasize the design intent distinction. |
|
shalom (manager) 2007-09-18 02:49 |
I would be wary of writing, "The always_latch procedure is identical to the always_comb procedure," because of the difference in how tools should relate to their intents. Also note that there is one statement in 9.2.2.2.2 that does not appear in 9.2.2.2: "Statements in an always_comb shall not include those that block, have blocking timing or event controls, or forkâjoin statements." It really should appear in 9.2.2.2. Actually, I think a 5th-order sub-clause is really exaggerated. I would prefer to fold 9.2.2.2.1 and 9.2.2.2.2 into 9.2.2.2. |
|
shalom (manager) 2007-10-14 05:46 |
Cliff Cummings vote comments: SVDB 1468 Minor No I would vote yes with the friendly amendment of replacing the personal pronoun "they" with "tools": "... , tools may check and warn if the behavior ..." |
|
Thomas R Alsop (administrator) 2007-10-15 09:43 |
Friendly update changing "they" to "tools" done at Cliff's request. |
|
mmaidment (manager) 2007-10-22 00:00 |
On October 15, 2007 the SV-BC unanimously approved the attached proposal. |
|
Neil Korpusik (administrator) 2007-11-21 07:58 |
The Champions approved the proposal in the Nov 8 conference call (with 1 abstain - Stu didn't get a chance to review ahead of time). |
|
Neil Korpusik (administrator) 2007-11-21 07:59 |
The proposal was unanimously approved by the Working Group in the November 15th, 2007 conference call. |
|
shalom (manager) 2008-05-30 08:21 |
The last "always_latch" in the proposal should be bold Courier. |
|
Stuart Sutherland (manager) 2008-06-03 01:05 |
The font error noted in bug note 6918 was corrected in Draft 6. |
| Mantis 1.1.7[^] Copyright © 2000 - 2008 Mantis Group |