Minutes, IBIS Quality Task Group 1 February 2022 12:00-13:00 EST (09:00-10:00 PST) ROLL CALL ANSYS Curtis Clark Intel Technology * Michael Mirmak Micron Technology * Randy Wolff Siemens EDA * Weston Beal Signal Integrity Software: * Mike LaBonte Teraspeed Labs: * Bob Ross Zuken USA: * Lance Wang Everyone in attendance marked by * NOTE: "AR" = Action Required. -----------------------MINUTES --------------------------- Mike LaBonte conducted the meeting. Call for IBIS related patent disclosures: - None Call for opens: - None Review of previous meeting minutes: Minutes from the January 25, 2021 meeting were reviewed. Bob Ross moved to accept the minutes. Lance Wang seconded the motion. Without objection, the minutes were approved. ARs: - AR: Mike LaBonte to send updated draft IBISCHK7.1.0 User Guide for review Not done. - AR: Mike LaBonte to investigate CMODEL_ERR_1 E,B confusion Done. - AR: Bob Ross to update BUG228 Done. - AR: Weston Beal to prepare presentation on IBIS Quality Specification and checklist Not done. NEW ITEMS: IBISCHK710 development: Mike LaBonte said he found from code inspection that CMODEL_ERR_1 is called in two places, once as bug and once as error. In both cases the problem was a memory allocation failure, so they should both generate bug messages. Bob Ross said a bug was an unknown error, something wrong in the code. Mike said we call it a bug, but in the current case it really a system problem, running out of memory. He said it would be hard to make a testcase fail. Bob asked Mike to writ ea bug report. Mike said he would, but it would not have a testcase. Bob agreed. AR: Mike LaBonte to write bug for CMODEL_ERR_1 Parser bug reports: BUG228: Bob Ross showed draft 3 of the bug. He had reduced the testcase size. Bob said the .ibs file could be checked as [IBIS Ver] 7.0, not 7.1, and the same two errors would appear. Randy Wolff agreed that the errors did appear in 7.0. He said there was a question as to which version to check .ibs files referenced from a .emd against, especially in the future when the latest version of IBIS increases beyond 7.1. The new draft had suggested wording for a message to explain why the .ibs file was being checked against a newer standard. Mike LaBonte felt it would be acceptable to check against IBIS 7.0 for now, and bump that up in the future as needed. Randy said it could be confusing to have yet another version number be used, one that is not the version of the .emd or .ibs file. Bob said IBIS 7.0 was the minimum required for the .ibs file to be used with .emd. Mike felt the word "upgraded" in Bob's suggested message implied that IBISCHK had modified the file. Randy said the message drafted in the previous meeting could have two sentences merged with a semicolon separator. Mike felt the <7.1> version number at the end of the sentence was not needed. Randy said we should require 7.0 for now, but in the future we might need to increase that requirement. We agreed on the message: Checking bug228.ibs for IBIS 7.0 Compatibility... The file has [IBIS Ver] <5.1>; however, IBIS 7.0 compatibility is required for IBIS files referenced from EMD. We agreed that the first line, which was what the current parser printed, should remain unchanged. The version in angle brackets would depend on the file being checked, but the angle brackets not not be printed. Randy said [Pin Mapping], which was related to the message listed in the bug, appeared in BIRD5.4. Weston Beal said all that mattered was the version of the file itself. AR: Bob Ross to update BUG228 with meeting feedback changes BUG229: Bob Ross showed the bug. He asked if it was a Windows bug, not ALL. Mike LaBonte agreed, noting the platform should also be PC. Bob made those changes. Bob said the expansion of the acronym UNC should be added, Universal Naming Convention. Bob said the bug would happen when the .pkg file was on a UNC path. Mike said it would happen when the .ibs file was on a UNC path. He said the reason the .pkg file was not found was not because it was on a UNC path, but rather because the dir command looked for it in C:\Windows. Weston Beal asked if the testcase might work in Windows PowerShell. Mike had not tried that, but he believed it would. He noted that the bug already had one workaround, but PowerShell might give us a second. Bob said EDA engineers should know how to use the workaround. Bob proposed classifying the bug as Annoying, Low. Mike agreed. Bob said we did not have any other Weston said it should check every .pkg file to search for the named [Package Model]. Bob said it should stop when it finds the first match. Mike suggested having the new code inspect all .pkg files as Weston had described. He noted that on Linux "ls -ld" was used, which gave a long listing with file attributes. On Windows "dir /b /l /a:-s-h-d" was used. Weston said the dir /l option forced lowercase output and /b gave a bare listing with only file names. The /a option selected only files, no directories, hidden, or system files. BUG230: Michael Mirmak said he would be raising a clock pin issue related to selectors in the IBIS ATM Task Group. He assumed from reading the specification that [Model Selector] and [Clock Pin] should interact with no problem, but IBISCHK had trouble with that. Mike asked if theer would be a bug report. He also noted that the bug report template had no place for a summary line. Michael said he might propose and updated template. IBIS quality checklist: Weston Beal said he did not yet have a presentation to give. He noted that the IBIS Quality Specification defined an IQ level 4 for power aware models, suggesting that with the power aware features in IBIS 5.2, that time was now. Tabled topics (no discussion without motion): - BIRD181.2 - IBISCHK security fixes Randy Wolff moved moved to adjourn. Michael Mirmak seconded. Without objection the meeting ended. Meeting ended: 13:07 ET Next meeting February 8, 2022