coreboot March 2015

coreboot@coreboot.org

62 participants
166 discussions

New Defects reported by Coverity Scan for coreboot
by scan-admin＠coverity.com 06 Mar '15

06 Mar '15

Hi, Please find the latest report on new defect(s) introduced to coreboot found with Coverity Scan. 30 new defect(s) introduced to coreboot found with Coverity Scan. 5 defect(s), reported by Coverity Scan earlier, were marked fixed in the recent build analyzed by Coverity Scan. New defect(s) Reported-by: Coverity Scan Showing 20 of 30 defect(s) ** CID 1287080: (FORWARD_NULL) /src/vendorcode/amd/agesa/f14/Proc/CPU/S3.c: 542 in SaveConditionalMsrDevice() /src/vendorcode/amd/agesa/f12/Proc/CPU/S3.c: 548 in SaveConditionalMsrDevice() /src/vendorcode/amd/agesa/f15/Proc/CPU/S3.c: 548 in SaveConditionalMsrDevice() /src/vendorcode/amd/agesa/f15tn/Proc/CPU/S3.c: 547 in SaveConditionalMsrDevice() /src/vendorcode/amd/agesa/f16kb/Proc/CPU/S3.c: 547 in SaveConditionalMsrDevice() ________________________________________________________________________________________________________ *** CID 1287080: (FORWARD_NULL) /src/vendorcode/amd/agesa/f14/Proc/CPU/S3.c: 542 in SaveConditionalMsrDevice() 536 if (CallPoint == INIT_RESUME) { 537 MemFS3GetCMsrDeviceRegisterList (Device, &RegisterHdr, StdHeader); 538 } else { 539 S3GetCMsrDeviceRegisterList (Device, &RegisterHdr, StdHeader); 540 } 541 >>> CID 1287080: (FORWARD_NULL) >>> Dereferencing null pointer "RegisterHdr". 542 for (i = 0; i < RegisterHdr->NumRegisters; i++) { 543 if (((Device->Mask1 & RegisterHdr->RegisterList[i].Mask1) != 0) && 544 ((Device->Mask2 & RegisterHdr->RegisterList[i].Mask2) != 0)) { 545 if (RegisterHdr->RegisterList[i].Type.SpecialCaseFlag == 0) { 546 LibAmdMsrRead (RegisterHdr->RegisterList[i].Address, (UINT64 *) *OrMask, StdHeader); 547 } else { /src/vendorcode/amd/agesa/f12/Proc/CPU/S3.c: 548 in SaveConditionalMsrDevice() 542 if (CallPoint == INIT_RESUME) { 543 MemFS3GetCMsrDeviceRegisterList (Device, &RegisterHdr, StdHeader); 544 } else { 545 S3GetCMsrDeviceRegisterList (Device, &RegisterHdr, StdHeader); 546 } 547 >>> CID 1287080: (FORWARD_NULL) >>> Dereferencing null pointer "RegisterHdr". 548 for (i = 0; i < RegisterHdr->NumRegisters; i++) { 549 if (((Device->Mask1 & RegisterHdr->RegisterList[i].Mask1) != 0) && 550 ((Device->Mask2 & RegisterHdr->RegisterList[i].Mask2) != 0)) { 551 if (RegisterHdr->RegisterList[i].Type.SpecialCaseFlag == 0) { 552 LibAmdMsrRead (RegisterHdr->RegisterList[i].Address, (UINT64 *) *OrMask, StdHeader); 553 } else { /src/vendorcode/amd/agesa/f15/Proc/CPU/S3.c: 548 in SaveConditionalMsrDevice() 542 if (CallPoint == INIT_RESUME) { 543 MemFS3GetCMsrDeviceRegisterList (Device, &RegisterHdr, StdHeader); 544 } else { 545 S3GetCMsrDeviceRegisterList (Device, &RegisterHdr, StdHeader); 546 } 547 >>> CID 1287080: (FORWARD_NULL) >>> Dereferencing null pointer "RegisterHdr". 548 for (i = 0; i < RegisterHdr->NumRegisters; i++) { 549 if (((Device->Mask1 & RegisterHdr->RegisterList[i].Mask1) != 0) && 550 ((Device->Mask2 & RegisterHdr->RegisterList[i].Mask2) != 0)) { 551 if (RegisterHdr->RegisterList[i].Type.SpecialCaseFlag == 0) { 552 LibAmdMsrRead (RegisterHdr->RegisterList[i].Address, (UINT64 *) *OrMask, StdHeader); 553 } else { /src/vendorcode/amd/agesa/f15tn/Proc/CPU/S3.c: 547 in SaveConditionalMsrDevice() 541 if (CallPoint == INIT_RESUME) { 542 MemFS3GetCMsrDeviceRegisterList (Device, &RegisterHdr, StdHeader); 543 } else { 544 S3GetCMsrDeviceRegisterList (Device, &RegisterHdr, StdHeader); 545 } 546 >>> CID 1287080: (FORWARD_NULL) >>> Dereferencing null pointer "RegisterHdr". 547 for (i = 0; i < RegisterHdr->NumRegisters; i++) { 548 if (((Device->Mask1 & RegisterHdr->RegisterList[i].Mask1) != 0) && 549 ((Device->Mask2 & RegisterHdr->RegisterList[i].Mask2) != 0)) { 550 if (RegisterHdr->RegisterList[i].Type.SpecialCaseFlag == 0) { 551 LibAmdMsrRead (RegisterHdr->RegisterList[i].Address, (UINT64 *) *OrMask, StdHeader); 552 } else { /src/vendorcode/amd/agesa/f16kb/Proc/CPU/S3.c: 547 in SaveConditionalMsrDevice() 541 if (CallPoint == INIT_RESUME) { 542 MemFS3GetCMsrDeviceRegisterList (Device, &RegisterHdr, StdHeader); 543 } else { 544 S3GetCMsrDeviceRegisterList (Device, &RegisterHdr, StdHeader); 545 } 546 >>> CID 1287080: (FORWARD_NULL) >>> Dereferencing null pointer "RegisterHdr". 547 for (i = 0; i < RegisterHdr->NumRegisters; i++) { 548 if (((Device->Mask1 & RegisterHdr->RegisterList[i].Mask1) != 0) && 549 ((Device->Mask2 & RegisterHdr->RegisterList[i].Mask2) != 0)) { 550 if (RegisterHdr->RegisterList[i].Type.SpecialCaseFlag == 0) { 551 LibAmdMsrRead (RegisterHdr->RegisterList[i].Address, (UINT64 *) *OrMask, StdHeader); 552 } else { ________________________________________________________________________________________________________ To view the defects in Coverity Scan visit, https://scan.coverity.com/projects/1016?tab=overview To manage Coverity Scan email notifications for "coreboot(a)coreboot.org", click https://scan.coverity.com/subscriptions/edit?email=coreboot%40coreboot.org&… .

1 0

[GSoC-2015] Interested in Participating - Discussing Ideas
by Varad Gautam 06 Mar '15

06 Mar '15

Hi, I would like to participate in Google Summer of Code 2015 with coreboot to work on one of the following ideas: - Enable coreboot + Tianocore over an ARM board: The Ideas page mentions something similar [1], but I think the hardware details are to be finalized. I worked on adding BeagleBone Black support to Tianocore during GSoC-2014 with Linaro, and so have some experience with JTAG, ARM bringup and bootloaders. Which mainboard would be the intended target? I have a HiKey board (HiSilicon Kirin 6220 Cortex-A53 ARM64) [2] which could be used - I'm checking if coreboot already supports a similar SoC to base the port on. - Eliminate Video BIOS requirement for AMD/ATI graphics init: I understand this would involve dumping AtomBIOS ROM and performing those initializations under coreboot. I came across atomdis [3] and am looking for a Video BIOS ROM I could try disassembling. I need some input on how I could go about the implementation of this idea - what hardware would be suitable / what the boot flow would be. Although I haven't worked with the Linux graphics stack yet, I would like to get my hands dirty. I am currently looking through the resources at [4]. I recently tried running coreboot + SeaBIOS + SYSLINUX under QEMU (as suggested on the Easy tasks page [5]), and it seems to work alright - but is it possible to find a SYSLINUX ELF image so I can eliminate SeaBIOS? I would be glad to have your suggestions on how I could get started. [1] http://www.coreboot.org/Project_Ideas#coreboot_ARM_SoC.27s_mainboard_port [2] https://www.96boards.org/products/hikey/ [3] http://www.phoronix.com/scan.php?page=article&item=amd_atombios_dumper&num=1 [4] http://www.coreboot.org/Project_Ideas#Native_graphics_init [5] http://www.coreboot.org/Easy_projects#Add.2Ftest_new_supported_payloads Thanks, Varad

2 1

coreboot GSoC 2015
by Marc Jones 05 Mar '15

05 Mar '15

coreboot has been accepted as a GSoC project . Woot! Please see the following announcement: http://blogs.coreboot.org/blog/2015/03/02/coreboot-gsoc-2015/ Interested students and mentors should visit the coreboot GSoC page: http://www.coreboot.org/GSoC Regards, Marc

1 0

Re: [coreboot] Long license headers (ANGER WARNING)
by ron minnich 04 Mar '15

04 Mar '15

> > Does this count as common sense? http://review.coreboot.org/#/c/8602/

1 0

Re: [coreboot] Long license headers (ANGER WARNING)
by Julius Werner 04 Mar '15

04 Mar '15

> Section 1 of the GPLv2 contains: "keep intact all the notices that refer to this License". Yes, but that only applies to verbatim copies. Section 2 explicitly allows you to distribute modified versions (and isn't incorporation into a larger code body always a modification?) as long as the whole is licensed under the terms of the GPL. The GPLv3 is much clearer on this, saying 'The work must carry prominent notices stating that it is released under this License [...]. This requirement modifies the requirement in section 4 [the equivalent to section 1 in v2] to “keep intact all notices”.' about modified works. Unfortunately we are using v2, but I'm assuming this is just a clarification of the same terms. I mean, just from a common sense perspective (although I know we are dealing with legal issues...), we all agree that it's okay to copy a single function from one GPL file to another without copying the whole license header, right (because we're doing that all the time)? So where is the difference between that and making a new file with your own (GPL-compliant) header and copying every function from another file into there? The copyright is still owned by the original author, but the license he grants you for using that code allows you to reuse and modify it however you like, including removing his original copyright notice (as long as you still adhere to the general attribution and licensing requirements of the GPL). (FWIW I think we should keep lines of the form "Copyright <year> <author>" intact anyway, but just as a courtesy, not because we're legally obligated. An the big blob with the FSF street address below can definitely be changed.)

1 0

Re: [coreboot] Long license headers (ANGER WARNING)
by ron minnich 04 Mar '15

04 Mar '15

> > http://review.coreboot.org/#/c/8602/ > If this works I'll update the wiki. ron

1 0

Re: [coreboot] Long license headers (ANGER WARNING)
by Patrick Georgi 04 Mar '15

04 Mar '15

2015-03-04 18:32 GMT+01:00 ron minnich <rminnich(a)gmail.com>: > Peter, per Patrick's note here, and discussions I've had with people who > actually are real lawyers, who earn their money doing IP licensing, I think > you are quite wrong. We should not change the license text in the files > *unless it is done in consultation with the copyright holders*. There are things we can do, before "perfect" kills "good" again: Decide how we want it done in the future, and formalize that somewhere. Then let existing copyright holders decide if they want to go for the new format (they're easy to find: they're listed right in the headers we're debating), and let them do it (they are allowed to, after all). Patrick -- Google Germany GmbH ABC-Str. 19 20354 Hamburg Registergericht und -nummer: Hamburg, HRB 86891 Sitz der Gesellschaft: Hamburg Geschäftsführer: Graham Law, Christine Elizabeth Flores

1 0

Re: [coreboot] Long license headers (ANGER WARNING)
by ron minnich 04 Mar '15

04 Mar '15

On Wed, Mar 4, 2015 at 6:10 AM Patrick Georgi via coreboot < coreboot(a)coreboot.org> wrote: > 2015-03-04 14:58 GMT+01:00 Peter Stuge <peter(a)stuge.se>: > > I haven't gotten the impression that anyone is suggesting to remove > > or even change any copyright notices. > > > > Copyright documents authorship (like git does too) but says nothing > > about licensing terms. > Section 1 of the GPLv2 contains: "keep intact all the notices that > refer to this License". > I'm not a lawyer or I'd have more interesting things to do than to > argue on mailing lists, but that doesn't look quite as clear cut to > me. > > Peter, per Patrick's note here, and discussions I've had with people who actually are real lawyers, who earn their money doing IP licensing, I think you are quite wrong. We should not change the license text in the files *unless it is done in consultation with the copyright holders*. The DoD, DOE, and other US Gov't entities are very particular about the copyright text as it appears in the file, and that it not be changed. Same with companies I've dealt with, including Google. If, however, you are secretly a lawyer too, (which would not surprise me given your broad skillset), then let us know :-) Also, of course people proposed changing the license text, per Paul's earlier comment: "In the past Ron also disagreed on changing headers by Samsung – taken from U-Boot I think – saying their lawyers drafted those and therefore it should stay that way." ron

1 0

Re: [coreboot] Long license headers (ANGER WARNING)
by Peter Stuge 04 Mar '15

04 Mar '15

ron minnich wrote: > > IANAL (I'm not giving legal advice, yada yada), but I don't think that is > > how this works. Companies you don't work for can't force you to do X or > > prevent you to do Y with their files. > > Copyright statements are pretty complicated, and I am not sure you are > right. I am afraid I can't take your advice ;-) > > You can change the code. But the copyright statement is pretty clear > that you can't remove the copyright. Copyright is one thing, licensing is a very different thing. I haven't gotten the impression that anyone is suggesting to remove or even change any copyright notices. Copyright documents authorship (like git does too) but says nothing about licensing terms. The discussion is about having a uniform and short way to express the license terms for individual files. > Consider the simplest case: change the copyright to say V3 instead of V2. > Would that be acceptable? You are confusing copyright with licensing here. Copyright is a year and an author. Those do not change. Licensing terms can be changed with permission of the author. But if an author contributed code licensed as GPLv2 then the license can not be changed to v3 without the author's permission. How to document such permission is a different discussion. However, if the author has contributed code licensed as GPLv2 then anyone who has that code can do, at a minimum, everything that the license permits them to do. Since the license header text is also covered by the license, replacing it with an equivalent seems completely unproblematic to me. > Generally you're supposed to take the text in the copyright as given and > not change it. Certainly older coreboot code with US Gov't copyright made > that clear. Copyright yes of course. Licensing is orthogonal. > I've had it stated pretty clearly to me by people who are lawyers that > it's a very bad idea to start rearranging copyright statements. If they are the source of your confusion between copyright and licensing they probably shouldn't be giving advice on this topic anyway. //Peter

2 1

Re: [coreboot] Long license headers (ANGER WARNING)
by ron minnich 04 Mar '15

04 Mar '15

On Tue, Mar 3, 2015 at 7:32 PM Julius Werner <jwerner(a)chromium.org> wrote: > > > IANAL (I'm not giving legal advice, yada yada), but I don't think that is > how this works. Companies you don't work for can't force you to do X or > prevent you to do Y with their files. > Copyright statements are pretty complicated, and I am not sure you are right. I am afraid I can't take your advice ;-) You can change the code. But the copyright statement is pretty clear that you can't remove the copyright. And, past a point, you can't change the copyright, or it's meaningless. Consider the simplest case: change the copyright to say V3 instead of V2. Would that be acceptable? Generally you're supposed to take the text in the copyright as given and not change it. Certainly older coreboot code with US Gov't copyright made that clear. I've had it stated pretty clearly to me by people who are lawyers that it's a very bad idea to start rearranging copyright statements. At the very least, it's going to upset people you don't want to upset; why do that? ron

1 0

Jump to page:

2024

2023

2022

2021

2020

2019

2018

2017

2016

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

coreboot March 2015