Sbis visual c runtime error

sbis visual c runtime error

python, c++ and rust, particularly those with garbage-collection. on [3] is to export SBI as a library, sort of how the vDSO works on. When you try to install an update for the.NET Framework 1.0, 1.1, 2.0, 3.0, or 3.5, you may receive Windows Update error code "0x643" or Windows Installer. The file Trojans-001.sbi-20140115.cab is missing or corrupt. Microsoft Visual C++ Runtime Library. Runtime Error! Program: C:\.

youtube video

Microsoft Visual C++ Runtime Library Error SOLUCION Windows 10 11

Sbis visual c runtime error - can

SBI - taking a step back

Alex Bradbury's profile photo

Alex Bradbury

unread,
Nov 11, 2018, 1:21:50 AM11/11/18

Reply to author

Sign in to reply to author

Forward

Sign in to forward

Delete

You do not have permission to delete messages in this group

Link

Report message as abuse

Sign in to report message as abuse

Show original message

Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message

to [email protected]

I've been following the recent discussions on SBI extensions with
great interest. Firstly, thank you to everyone who has been
contributing SBI proposals and documenting the current de facto SBI
standard.

I do have one request - could we all take a step back and try to
define the role and driving principles of the SBI? What problems
should it be solving? What problems shouldn't be solved via the SBI? I
understand once the framework is in place there's scope for people to
use it in a wide variety of ways, but it's really hard to judge the
various proposals without any clear definition of the SBI
requirements. Proposing incremental ABI additions seems like the wrong
way to go about development without agreeing this first.

The nearest thing we have to the guiding principles of the SBI is the
stated aim in the current privspec draft "Using a single SBI across
all SEE implementations allows a single OS binary image to run on any
SEE". The document currently proposes SBI calls for:
* Initialising, saving, and restoring user-mode state outside of the
standard FDQ extensions
* Inter-processor interrupts
* Clear timer interrupts
* Mask, unmask, and query the cause of external interrupts
(apparently, all access to the PLIC should be through the SBI?)
* Modifying the time/cycle/instret CSRs
* Scheduling timer interrupts
* TLB shootdown

This is quite an expansion over the current SBI (which seems to offer
a rather arbitrary subset of platform-specific functionality). I would
observe that none of the proposals to extend the SBI so far actually
cover all of these cases. Is that because there is disagreement over
whether the SBI should serve all these purposes, or are the proposals
simply incomplete?

Based on a recent software working group call, it seems there is a
desire for the SBI to ensure that a single kernel works across all
RISC-V implementations, including support for custom extensions (e.g.
save/restore of user-visible state as mentioned above). I'm concerned
that reaching this goal will require an increasing number of hooks
into the kernel effectively making the SBI a mini-OS. e.g. for the
tagged memory mechanism in lowRISC, saving/restoring the tagctrl
register is a necessary component of kernel support but it's also
necessary to modify the virtual memory system to support (at a
minimum) clearing tags on pages and propagating tags upon
copy-on-write.

Other questions/concerns/thoughts:
* Is there actually consensus amongst developers of S-mode code
(kernels, bootloaders etc) that the SBI is a desirable abstraction?
* Shouldn't there be an interface to allow S-mode code to control
delegation from M-mode (the SBI/SEE)? e.g. if my OS would prefer to
supply its own floating point emulation code. Or maybe an alternative
mechanism to swap out the M-mode handler code (as Ron has been
advocating)
* We currently have 3 base ISAs - RV32I, RV32E, RV64I. How is the SBI
intended to work in the case where a hart supports switching between
these? Surely an S-mode OS may be running with a different ISA variant
to the M-mode SBI/SEE? How is this handled?
* There are ideas floating around about SBI calls being more like
function calls (rather than the current ecall mechanism). Has anyone
written down how this would work?

How about we start by agreeing and enumerating the requirements, then
work our way up?

Thanks,

Alex

Bin Meng's profile photo

Bin Meng

unread,
Nov 11, 2018, 6:34:21 AM11/11/18

Reply to author

Sign in to reply to author

Forward

Sign in to forward

Delete

You do not have permission to delete messages in this group

Link

Report message as abuse

Sign in to report message as abuse

Show original message

Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message

to [email protected], [email protected]

Hi Alex,


On Sun, Nov 11, 2018 at 5:21 AM Alex Bradbury <[email protected]> wrote:
>
> I've been following the recent discussions on SBI extensions with
> great interest. Firstly, thank you to everyone who has been
> contributing SBI proposals and documenting the current de facto SBI
> standard.
>
> I do have one request - could we all take a step back and try to
> define the role and driving principles of the SBI? What problems
> should it be solving? What problems shouldn't be solved via the SBI? I

I agree. Maybe one more question: why is SBI a must-have for RISC-V?
To me current SBI is not absolutely necessary. All these
functionalities can be implemented by the S-mode OS. Why do they have
to be M-mode? One may say some hardware components can only be
accessed by M-mode (eg: IPI) but why should hardware expose such
limitation?


> understand once the framework is in place there's scope for people to
> use it in a wide variety of ways, but it's really hard to judge the
> various proposals without any clear definition of the SBI
> requirements. Proposing incremental ABI additions seems like the wrong
> way to go about development without agreeing this first.
>
> The nearest thing we have to the guiding principles of the SBI is the
> stated aim in the current privspec draft "Using a single SBI across
> all SEE implementations allows a single OS binary image to run on any
> SEE". The document currently proposes SBI calls for:

Shouldn't we clarify why such goal that "Using a single SBI across all

SEE implementations allows a single OS binary image to run on any SEE"

is absolutely necessary? Eventually we will move all platform-specific
stuff into the SBI implementation and that makes SBI another mini-OS
(just as you pointed out below).


> * Initialising, saving, and restoring user-mode state outside of the
> standard FDQ extensions
> * Inter-processor interrupts
> * Clear timer interrupts
> * Mask, unmask, and query the cause of external interrupts
> (apparently, all access to the PLIC should be through the SBI?)

I hope we will not go this way. See below.


> * Modifying the time/cycle/instret CSRs
> * Scheduling timer interrupts
> * TLB shootdown
>

Basically I think RISC-V should allow S-mode OS to handle things like
timer interrupts, IPI, etc, to reduce interrupt latency. I hope S-mode
can directly handle these hardware components instead of going through
the M-mode firmware. There is needs for real-time performance and
determinism.


> This is quite an expansion over the current SBI (which seems to offer
> a rather arbitrary subset of platform-specific functionality). I would
> observe that none of the proposals to extend the SBI so far actually
> cover all of these cases. Is that because there is disagreement over
> whether the SBI should serve all these purposes, or are the proposals
> simply incomplete?
>
> Based on a recent software working group call, it seems there is a
> desire for the SBI to ensure that a single kernel works across all
> RISC-V implementations, including support for custom extensions (e.g.
> save/restore of user-visible state as mentioned above). I'm concerned
> that reaching this goal will require an increasing number of hooks
> into the kernel effectively making the SBI a mini-OS. e.g. for the
> tagged memory mechanism in lowRISC, saving/restoring the tagctrl
> register is a necessary component of kernel support but it's also
> necessary to modify the virtual memory system to support (at a
> minimum) clearing tags on pages and propagating tags upon
> copy-on-write.
>
> Other questions/concerns/thoughts:
> * Is there actually consensus amongst developers of S-mode code
> (kernels, bootloaders etc) that the SBI is a desirable abstraction?

I don't think so :)


> * Shouldn't there be an interface to allow S-mode code to control
> delegation from M-mode (the SBI/SEE)? e.g. if my OS would prefer to
> supply its own floating point emulation code. Or maybe an alternative
> mechanism to swap out the M-mode handler code (as Ron has been
> advocating)

Agreed. We should allow some flexibility.


> * We currently have 3 base ISAs - RV32I, RV32E, RV64I. How is the SBI
> intended to work in the case where a hart supports switching between
> these? Surely an S-mode OS may be running with a different ISA variant
> to the M-mode SBI/SEE? How is this handled?

Do you mean switch between 32-bit and 64-bit as defined in the draft
priv spec v1.11?


> * There are ideas floating around about SBI calls being more like
> function calls (rather than the current ecall mechanism). Has anyone
> written down how this would work?
>
> How about we start by agreeing and enumerating the requirements, then
> work our way up?

Regards,
Bin
Luke Kenneth Casson Leighton's profile photo

Luke Kenneth Casson Leighton

unread,
Nov 11, 2018, 6:42:07 AM11/11/18

Reply to author

Sign in to reply to author

Forward

Sign in to forward

Delete

You do not have permission to delete messages in this group

Link

Report message as abuse

Sign in to report message as abuse

Show original message

Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message

to Alex Bradbury, [email protected]

On Sat, Nov 10, 2018 at 9:21 PM Alex Bradbury <[email protected]> wrote:

> requirements. Proposing incremental ABI additions seems like the wrong
> way to go about development without agreeing this first.

hmmm, indeed.


> The nearest thing we have to the guiding principles of the SBI is the
> stated aim in the current privspec draft "Using a single SBI across
> all SEE implementations allows a single OS binary image to run on any
> SEE".

section 1.1, V1.11-draft of privspec, SEE="Supervisor Execution Environment"


> The document currently proposes SBI calls for:

_whoa_ whoa, whoa, that's an alarmingly-large set of functionality to
suddenly encounter.



> This is quite an expansion over the current SBI (which seems to offer
> a rather arbitrary subset of platform-specific functionality).

with zero access to critical information behind the unethical
ITU-style closed doors of the RISC-V Foundation, basically from the
sparse and paltry information here we need to reverse-engineer the
intent.

"...... Using a single SBI across all SEE implementations allows a
single OS binary
image to run on any SEE. The SEE can be a simple boot loader and
BIOS-style IO system in a
low-end hardware platform, or a hypervisor-provided virtual machine in
a high-end server, or a
thin translation layer over a host operating system in an architecture
simulation environment."

from that statement, and from the list that you gave, alex, it looks
to me that the intent here is to re-create XEN.

without any kind of wider consultation.


> I would
> observe that none of the proposals to extend the SBI so far actually
> cover all of these cases. Is that because there is disagreement over
> whether the SBI should serve all these purposes, or are the proposals
> simply incomplete?

who knows? certain people do, and they're lobbing stuff over the
fence, ramming it down everyone's throats in a fait-accomplit fashion
and then claiming it's "open".

which is basically the definition of unethical behaviour.


> Based on a recent software working group call, it seems there is a
> desire for the SBI to ensure that a single kernel works across all
> RISC-V implementations, including support for custom extensions (e.g.
> save/restore of user-visible state as mentioned above). I'm concerned
> that reaching this goal will require an increasing number of hooks
> into the kernel effectively making the SBI a mini-OS.

likewise. the clue is in the "recursive" requirement: this looks
very much like a XEN style hypervisor.

"The ABI, SBI, and HBI are still a work-in-progress, but we are now
prioritizing support for
Type-2 hypervisors where the SBI is provided recursively by an S-mode OS."

where is _that_ work and discussion being published??

i'm getting really alarmed that the RISC-V Foundation is operating
extremely unethcially, by isolating everyone from discussion and
forcing the rest of the community to have to work out what the hell is
going on... by the time things are actually released it's far, far too
late,

much worse, anyone OUTSIDE of this unethically-developed process who
submits upstream patches which adds anything that they actually *want*
is quite likely to have their patches REJECTED out of hand because
they're "not submitted through the proper [unethical] channels".


> e.g. for the
> tagged memory mechanism in lowRISC, saving/restoring the tagctrl
> register is a necessary component of kernel support but it's also
> necessary to modify the virtual memory system to support (at a
> minimum) clearing tags on pages and propagating tags upon
> copy-on-write.

that's similar to what IIT Madras are doing: they have some
memory-tracking that helps detect invalid malloc / free, and i talked
to them about the idea of also adding ref-counters for languages like
python, c++ and rust, particularly those with garbage-collection.



> Other questions/concerns/thoughts:
> * Is there actually consensus amongst developers of S-mode code
> (kernels, bootloaders etc) that the SBI is a desirable abstraction?

if it was being properly publicly discussed in an OPEN fashion, and
the code being PROPERLY developed as an organic project, i would not
be concerned.

as it is, we know that the RISC-V Foundation has a consistent track
record of operating completely unethically by shoving code at people.

do we *really* want something as mission-critical as a full
Hypervisor to be developed and then controlled by a known-unethical
organisation, running on our machines?

will we be told "well if you don't like it you can always fork it",
when we *know* that's totally financially unrealistic for small
organisations and companies to even contemplate.


> How about we start by agreeing and enumerating the requirements, then
> work our way up?

that would be a good idea.

l.
Alex Bradbury's profile photo

Alex Bradbury

unread,
Nov 11, 2018, 8:11:14 AM11/11/18

Reply to author

Sign in to reply to author

Forward

Sign in to forward

Delete

You do not have permission to delete messages in this group

Link

Report message as abuse

Sign in to report message as abuse

Show original message

Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message

to Luke Kenneth Casson Leighton, [email protected]

Hi Luke. I appreciate your frustration but I'd really like to keep
this thread focused on the specific technical topics I raised.

You've pointed out previously that the RISC-V mailing lists are often
not very welcoming. I think you should consider:
1) The individuals working on standards within the RISC-V Foundation
processes are doing so with good intent, doing the best they can with
the resources they have.
2) Criticism of this intensity, even when directed at "the Foundation"
rather than personally might put such individuals off communicating on
the public lists.

Thanks,

Alex
Luke Kenneth Casson Leighton's profile photo

Luke Kenneth Casson Leighton

unread,
Nov 11, 2018, 10:46:00 AM11/11/18

Reply to author

Sign in to reply to author

Forward

Sign in to forward

Delete

You do not have permission to delete messages in this group

Link

Report message as abuse

Sign in to report message as abuse

Show original message

Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message

to Alex Bradbury, [email protected]

---
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68
i'd like to see that happen as well, in an open fashion, where
everyone is invited to participate, and feels welcome. it would be
absolutely fantastic, to be able to have up-to-date information and
access to the resources and discussions that allow everyone to
participate.

except that's not the case, is it? and denying it and making excuses
is not going to help the situation get any better, is it? and
expecting people to WASTE THEIR TIME *reverse-engineering* private
closed discussions... and then because it's flat-out impossible to get
it right when the information and context isn't avaiable, i have
bullies fucking well REPEATEDLY poisoning discussions by saying
"you're wrong, you're wrong, but you don't know that because WE talked
about it on a PRIVATE list that YOU don't have access to, because
you're such an oh-so-high-and-mighty ethical little tosspot, har har
we had a real laugh about that one, at your expense, loserr"


right now there's dozens of people over the years - long before i got
involved - who have been made to feel so unwelcome that they have
completely given up. one organisation is even looking to set up a
complete fork of the entire RISC-V infrastructure, they're so pissed
off.

i mean, for f***'s sake, it's right here, the words used by the
director, rick o'connor, in the barcelona workshop announcement:

" Hear from from the major players in the RISC-V ecosystem, including
the leading technology companies and research institutions driving the
RISC-V ISA specification"

hear from the companies and the research institutes, give them credit
where it's due... oh and anyone involved in the libre and open world
who contributed and put hard work into making RISC-V a success can go
f*** themselves, because we can't be bothered to mention them, at all.

the systemic laws of organisations are systemic LAWS for a reason,
and failing to acknowledge credit for peoples' efforts is a systemic
law violation, plain and simple.




> You've pointed out previously that the RISC-V mailing lists are often
> not very welcoming. I think you should consider:
> 1) The individuals working on standards within the RISC-V Foundation
> processes are doing so with good intent, doing the best they can with
> the resources they have.

repeatedly telling me and others who have completely given up that
"they're worthless", or "they're wrong" or "they're wasting
everybody's time" or failing to acknowledge their needs and
constructive input is not even *remotely* "good intent".

telling someone that their constructive criticism is an "ad hominem
attack" is not even remotely "good intent": it's a way to deflect and
deny the truth of an ongoing long-running damaging behavioural
pattern.

what's going on is called absolute arrogant power-abusive bullying,
plain and simple.



> 2) Criticism of this intensity, even when directed at "the Foundation"
> rather than personally might put such individuals off communicating on
> the public lists.

they'll continue to do what is useful to them. that includes
bullying. i've been targetted by bullies my entire life, for speaking
up when nobody else will, because they're too afraid or too vested to
do so.

i'll speak the truth, living to the integrity that i've set, and
nobody's going to tell me i can't.

when people do something good, i'll say so. when people do something
bad, i'll say so. you can't ask fairer than that.

where and when the RISC-V Foundation and the people behind it start
doing *GOOD*, i'll be their loudest, most vocal and most loyal
supporter. and you've already seen me do that: i've made it clear,
repeatedly, that i am deeply impressed and respectful of what they've
achieved.

however i cannot sit by and watch unethical behaviour unfold right in
front of my eyes, and i can NOT support unethical behaviour, or fund
it, or provide resources to organisations or individuals that operate
unethically, because to do so it itself unethical, and i absolutely
cannot do that.

it may sound weird to be both respectful of someone's accomplishments
*and* have some criticism of someone's behaviour: that's just how it
is. there's no conflict there, at all. and i say that because i've
seen people already twist my words and focus on the constructive
criticism.

you want me to stop? HELP get the unethical behaviour of the RISC-V
Foundation FIXED, by acknowledging it, discussing it, and recommending
alternatives and solutions.

if those alternatives and solutions are ethical, and properly adopted,
i'll become the RISC-V Foundation's most loyal supporter and advocate,
you can absolutely count on it.

l.
Bruce Hoult's profile photo

Bruce Hoult

unread,
Nov 11, 2018, 10:59:31 AM11/11/18

Reply to author

Sign in to reply to author

Forward

Sign in to forward

Delete

You do not have permission to delete messages in this group

Link

Report message as abuse

Sign in to report message as abuse

Show original message

Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message

to Luke Kenneth Casson Leighton, Alex Bradbury, [email protected]

Luke Kenneth Casson Leighton's profile photo

Luke Kenneth Casson Leighton

unread,
Nov 11, 2018, 11:47:33 AM11/11/18

Reply to author

Sign in to reply to author

Forward

Sign in to forward

Delete

You do not have permission to delete messages in this group

Link

Report message as abuse

Sign in to report message as abuse

Show original message

Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message

to Bruce Hoult, Alex Bradbury, [email protected]

On Sun, Nov 11, 2018 at 6:59 AM Bruce Hoult <[email protected]> wrote:

> Luke, Alex is a very reasonable person who is trying to help.

i know. he's acting with good intent and is following some of the
guidelines here:
http://www.crnhq.org/content.aspx?file=66138

Notice

How to obtain help and support for this security update


Help installing updates:
Support for Microsoft Update

Security solutions for IT professionals:
TechNet Security Troubleshooting and Support

Help protect your computer that is running Windows from viruses and malware:
Virus Solution and Security Center

Local support according to your country:
International Support

Symptoms of the problem

When you try to install an update for the Microsoft .NET Framework 1.0, for the Microsoft .NET Framework 1.1, the Microsoft .NET Framework 2.0, the Microsoft .NET Framework 3.0, or for the Microsoft .NET Framework 3.5, you may receive the following Microsoft Windows Update error code:

Or, you may receive the following Microsoft Windows Installer error code:

This problem may occur when you use certain system configurations.

Cause

Windows Update error code "0x643" and Windows Installer error code "1603" are generic error codes. Typically, these error codes mean that the update was not installed. These errors codes are usually caused by a corruption in the .NET Framework installation or by an inconsistency on the MSI database state.

Resolution

To resolve this problem, you must fix the MSI software update registration corruption issues, or uninstall multiple versions of the .NET Framework by using the .NET Framework Cleanup Tool, and then reinstall those components.

To troubleshoot this problem, follow these steps starting with Method 1. Proceed to the Method 2 if Method 1 does not resolve the problem.

Method 1: Fix MSI software update registration corruption issues

For more information about how to fix MSI software update registration corruption issues, click the following article number to view the article in the Microsoft Knowledge Base:

971187 How to fix MSI software update registration corruption issuesTo fix this problem automatically, click the Fix it button or link in the 971187 article.


Then visit the Windows Update Web site and try to install the updates:

http://update.microsoft.comIf the problem is not resolved, proceed to Method 2.

Method 2: Repair the .NET Framework

To repair the .NET Framework, you must uninstall multiple versions of the .NET Framework by using the .NET Framework Cleanup Tool, and then reinstall the components.


To do this, follow these steps:

  1. Click the following link to download the .NET Framework Cleanup Tool:

    http://blogs.msdn.com/b/astebner/archive/2008/08/28/8904493.aspx

  2. Use the .NET Framework Cleanup Tool to uninstall the .NET Framework versions 1.0, 1.1, 2.0, 3.0, and 3.5

    Note The cleanup tool does not let you remove the .NET Framework 2.0 in Windows Vista or later versions because the .NET Framework is installed as an operating system component.

  3. Restart the computer.

  4. Download and install the following components:

  5. Restart the computer.

  6. Visit the Windows Update Web site and install the updates:

    http://update.microsoft.com

STILL NEED HELP?

If this article does not answer your question, click this button to pose your question to other community members at Microsoft Community:

Ask a question


More Information

You may experience this problem when you use any of the following configurations:

  • Microsoft .NET Framework 1.0 Service Pack 3 when you use it with any of the following operating systems:

    • Microsoft Windows 2000 Advanced Server Service Pack 4 (SP4)

    • Microsoft Windows 2000 Datacenter Server SP4

    • Microsoft Windows 2000 Professional SP4

    • Microsoft Windows 2000 Server SP4

    • Windows Server 2003, Datacenter Edition for Itanium-based Systems SP1

    • Windows Server 2003, Datacenter Edition for Itanium-based Systems SP2

    • Windows Server 2003, Datacenter Edition SP1

    • Windows Server 2003, Datacenter Edition SP2

    • Windows Server 2003, Datacenter x64 Edition SP2

    • Windows Server 2003, Datacenter x64 Edition

    • Windows Server 2003, Enterprise Edition for Itanium-based Systems SP1

    • Windows Server 2003, Enterprise Edition for Itanium-based Systems SP2

    • Windows Server 2003, Enterprise Edition SP1

    • Windows Server 2003, Enterprise Edition SP2

    • Windows Server 2003, Enterprise x64 Edition SP2

    • Windows Server 2003, Enterprise x64 Edition

    • Windows Server 2003, Standard Edition 32-bit x86 SP1

    • Windows Server 2003, Standard Edition 32-bit x86 SP2

    • Windows Server 2003, Standard x64 Edition SP2

    • Windows Server 2003, Standard x64 Edition

    • Windows Server 2003, Web Edition SP1

    • Windows Server 2003, Web Edition SP2

    • Windows Vista Business

    • Windows Vista Enterprise

    • Windows Vista Home Basic

    • Windows Vista Home Premium

    • Windows Vista Ultimate

    • Windows XP Home Edition SP2

    • Windows XP Media Center Edition 2005

    • Windows XP Professional SP2

    • Windows XP Professional x64 Edition

    • Windows XP Professional x64 Edition SP2

    • Windows XP Tablet PC Edition SP2

  • Microsoft .NET Framework 1.1 Service Pack 1 when you use it with any of the following operating systems:

    • Windows 2000 Advanced Server SP4

    • Windows 2000 Datacenter Server SP4

    • Windows 2000 Professional SP4

    • Windows 2000 Server SP4

    • Windows Server 2003, Datacenter Edition for Itanium-based Systems SP1

    • Windows Server 2003, Datacenter Edition for Itanium-based Systems SP2

    • Windows Server 2003, Datacenter Edition SP2

    • Windows Server 2003, Datacenter x64 Edition SP2

    • Windows Server 2003, Datacenter x64 Edition

    • Windows Server 2003, Enterprise Edition for Itanium-based Systems SP1

    • Windows Server 2003, Enterprise Edition for Itanium-based Systems SP2

    • Windows Server 2003, Enterprise Edition SP2

    • Windows Server 2003, Enterprise x64 Edition SP2

    • Windows Server 2003, Enterprise x64 Edition

    • Windows Server 2003, Standard Edition 32-bit x86 SP2

    • Windows Server 2003, Standard x64 Edition SP2

    • Windows Server 2003, Standard x64 Edition

    • Windows Server 2003, Web Edition SP2

    • Windows Vista Business

    • Windows Vista Enterprise

    • Windows Vista Home Basic

    • Windows Vista Home Premium

    • Windows Vista Ultimate

    • Windows XP Home Edition SP2

    • Windows XP Media Center Edition 2005

    • Windows XP Professional SP2

    • Windows XP Professional x64 Edition

    • Windows XP Professional x64 Edition SP2

  • Microsoft .NET Framework 2 or Microsoft .NET Framework 2 SP1 when you use it with any of the following operating systems:

    • Windows 2000 Advanced Server SP4

    • Windows 2000 Datacenter Server SP4

    • Windows 2000 Professional SP4

    • Windows 2000 Server SP4

    • Windows Server 2003, Datacenter Edition for Itanium-based Systems SP1

    • Windows Server 2003, Datacenter Edition for Itanium-based Systems SP2

    • Windows Server 2003, Datacenter Edition SP1

    • Windows Server 2003, Datacenter Edition SP2

    • Windows Server 2003, Datacenter x64 Edition

    • Windows Server 2003, Datacenter x64 Edition SP2

    • Windows Server 2003, Enterprise Edition for Itanium-based Systems SP1

    • Windows Server 2003, Enterprise Edition for Itanium-based Systems SP2

    • Windows Server 2003, Enterprise Edition SP1

    • Windows Server 2003, Enterprise Edition SP2

    • Windows Server 2003, Enterprise x64 Edition

    • Windows Server 2003, Enterprise x64 Edition SP2

    • Windows Server 2003, Standard Edition 32-bit x86 SP1

    • Windows Server 2003, Standard Edition 32-bit x86 SP2

    • Windows Server 2003, Standard x64 Edition

    • Windows Server 2003, Standard x64 Edition SP2

    • Windows Server 2003, Web Edition SP1

    • Windows Server 2003, Web Edition SP2

    • Windows XP Home Edition SP2

    • Windows XP Media Center Edition 2005

    • Windows XP Professional SP2

    • Windows XP Professional x64 Edition

    • Windows XP Professional x64 Edition SP2

    • Windows XP Tablet PC Edition SP2

  • Microsoft .NET Framework 3.0 or Microsoft .NET Framework 3.0 SP1 when you use it with any of the following operating systems:

    • Windows Server 2003, Datacenter Edition for Itanium-based Systems SP1

    • Windows Server 2003, Datacenter Edition for Itanium-based Systems SP2

    • Windows Server 2003, Datacenter Edition SP1

    • Windows Server 2003, Datacenter Edition SP2

    • Windows Server 2003, Datacenter x64 Edition

    • Windows Server 2003, Datacenter x64 Edition SP2

    • Windows Server 2003, Enterprise Edition for Itanium-based Systems SP1

    • Windows Server 2003, Enterprise Edition for Itanium-based Systems SP2

    • Windows Server 2003, Enterprise Edition SP1

    • Windows Server 2003, Enterprise Edition SP2

    • Windows Server 2003, Enterprise x64 Edition

    • Windows Server 2003, Enterprise x64 Edition SP2

    • Windows Server 2003, Standard Edition 32-bit x86 SP1

    • Windows Server 2003, Standard Edition 32-bit x86 SP2

    • Windows Server 2003, Standard x64 Edition

    • Windows Server 2003, Standard x64 Edition SP2

    • Windows Server 2003, Web Edition SP1

    • Windows Server 2003, Web Edition SP2

    • Windows XP Home Edition SP2

    • Windows XP Media Center Edition 2005

    • Windows XP Professional SP2

    • Windows XP Professional x64 Edition

    • Windows XP Professional x64 Edition SP2

    • Windows XP Tablet PC Edition SP2

  • Microsoft .NET Framework 3.5 when you use it with any of the following operating systems:

    • Windows Server 2003, Datacenter Edition for Itanium-based Systems SP1

    • Windows Server 2003, Datacenter Edition for Itanium-based Systems SP2

    • Windows Server 2003, Datacenter Edition SP1

    • Windows Server 2003, Datacenter Edition SP2

    • Windows Server 2003, Datacenter x64 Edition

    • Windows Server 2003, Datacenter x64 Edition SP2

    • Windows Server 2003, Enterprise Edition for Itanium-based Systems SP1

    • Windows Server 2003, Enterprise Edition for Itanium-based Systems SP2

    • Windows Server 2003, Enterprise Edition SP1

    • Windows Server 2003, Enterprise Edition SP2

    • Windows Server 2003, Enterprise x64 Edition

    • Windows Server 2003, Enterprise x64 Edition SP2

    • Windows Server 2003, Standard Edition 32-bit x86 SP1

    • Windows Server 2003, Standard Edition 32-bit x86 SP2

    • Windows Server 2003, Standard x64 Edition

    • Windows Server 2003, Standard x64 Edition SP2

    • Windows Server 2003, Web Edition SP1

    • Windows Server 2003, Web Edition SP2

    • Windows Vista Business

    • Windows Vista Enterprise

    • Windows Vista Home Basic

    • Windows Vista Home Premium

    • Windows Vista Ultimate

    • Windows XP Home Edition SP2

    • Windows XP Media Center Edition 2005

    • Windows XP Professional SP2

    • Windows XP Professional x64 Edition

    • Windows XP Professional x64 Edition SP2

    • Windows XP Tablet PC Edition SP2

For more information about another error message that you may receive when you try to install this update, click the following article number to view the article in the Microsoft Knowledge Base:

923101Error message when you try to install a security update for the .NET Framework 2.0 on a computer that is running Windows Server 2003 x64 Edition: "Error 1324. The folder 'Program Files' contains an invalid character"

37449y#Empathy



> You're actually capable of staying within the bounds of technical discussion.

that's right - i am!


> I've seen it in the last week or so, in comp.arch, where you managed to reply
> to me in a civilised way, and even quote me approvingly from time to time.

that's right - i did. so you know it's perfectly possible. and, if
you've reviewed the libre-risc-dev (a low-traffic mailing list as it's
only just started), you'll see more evidence of that.

if you go further back than that, you'll find evidence that i've been
the technical lead on large software libre projects for over two
decades, with in some cases over 20 simultaneous part-time
contributors.

so the question to ask is not "how* come you are civilised and
capable of staying within the bounds of technical discussion on other
forums: the real question is, what's different about *these* forums?

that's the *real* question.


> Please try to do the same here in future.

i would absolutely love to see these forums be welcoming and
productive technical forums where *everyone* is welcome, where they
can get fully up-to-date discussions and information (so that effort
is not duplicated or wasted), and can be directly involved in the
development and direction of RISC-V, where they can express their
needs and requirements clearly and be listened to, and their
contributions are appreciated, respected and encouraged.

does that sound reasonable?

l.
Bruce Hoult's profile photo

Bruce Hoult

unread,
Nov 11, 2018, 12:00:14 PM11/11/18

Reply to author

Sign in to reply to author

Forward

Sign in to forward

Delete

You do not have permission to delete messages in this group

Link

Report message as abuse

Sign in to report message as abuse

Show original message

Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message

to Luke Kenneth Casson Leighton, Alex Bradbury, [email protected]

Luke Kenneth Casson Leighton's profile photo

Luke Kenneth Casson Leighton

unread,
Nov 11, 2018, 1:46:55 PM11/11/18

Reply to author

Sign in to reply to author

Forward

Sign in to forward

Delete

You do not have permission to delete messages in this group

Link

Report message as abuse

Sign in to report message as abuse

Show original message

Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message

to Bruce Hoult, Alex Bradbury, [email protected]

---
crowd-funded eco-conscious hardware: https://www.crowdsupply.com/eoma68

On Sun, Nov 11, 2018 at 8:00 AM Bruce Hoult <[email protected]> wrote:
>
> On Sat, Nov 10, 2018 at 11:46 PM, Luke Kenneth Casson Leighton <[email protected]> wrote:
>>
>> so the question to ask is not "how* come you are civilised and
>> capable of staying within the bounds of technical discussion on other
>> forums: the real question is, what's different about *these* forums?
>
>
> In the last few hours I've seen you being adversarial with not only
> Alex Bradbury,

if you look carefully, alex was trying to be reasonable, and play a
mediatorial role... except if you look carefully at the advice given
on crhnq, a mediator needs to respect *both* sides, and relay (reflect
back) the position of *both* parties.

regrettably, he didn't do that, and unfortunately, all it did was
make it look he wasn't listening to what i'm saying. as i was
writing, i was thinking, "damn, damn damn, i really like alex, how the
hell can i make this look any better, so it doesn't reflect badly on
him?" and, sadly, as he (you, hello, yes, sorry to be talking _about_
you, alex) hadn't summarised and acknowledged both perspectives as a
mediator normally would, there really wasn't a way to do that.

> but also Nick Kossifidis

if you look carefully at what nick's writing, he made good points
from the field in which he has experience.. unfortunately he laced it
with hyperbolic adversarial questions that were very carefully
designed to ridicule rather than critique not only the proposal that i
made, they were also very very carefully crafted to ridicule me, as a
person, as well.

you'll notice that he dodged the positive aspects that were supported
by other people's constructive input, and focussed on generating as
much negativity as he could, through the use of what can only be
termed "hysteria" questions, hoping to assign them to me (without my
consent) and use them rhetorically for the purposes of ridicule.

the initial (thinly-veiled) adversarial attacks that he launched were
sufficiently over the top that someone else had to step in, and hint
that he was going a little too far. nick made a conscious choice to
ignore those hints.

the question here i have to ask, is, can you see that that's what
nick did, and, if you can, why do you feel that it's acceptable
behaviour on a technical forum?


> (who you for unknown reason persist in addressing as "Mick"),

sorry, nick: small high-resolution screen here, and i run the
brightness way down and the room lights off, most of the time.

> and Tommy Murphy.

tommy, bless him, i know he's trying to help. you may be confusing
the pain that i am experiencing through having to deal with the
ongoing bullying and expressing that, and thus interpreting my
responses to him as "this must be an attack on tommy". categorically
and absolutely it is not.


> And adopting a derisory tone with others such as Ron Minnich.

what?? not at all! ron makes a lot of sense, and i've said so quite
clearly!! what on *earth* gave you the impression that i've been
"derisory" with ron?? i've made it quite clear that the points he's
made about being able to operate fully in M-Mode are good ones!


> There's a common thread here.

there is, indeed. the challenge is, then, if the opportunity to
attack arises, if someone appears to be vulnerable, in pain, or
appears to not be knowledgeable, is for people to step up to the plate
and *not* react in a derisory fashion, *not* attack or denigrate them,
make them feel so unwelcome and so despised that it actually
deteriorates their physical health.

you know i speak my mind, and i made a decision 20 years ago to speak
truth, even when it's uncomfortable truth. that's what people face
here, and i know they don't like it.

i told you: ethical's a bitch. truth, awareness, love and creativity
aren't "nice", they're damn uncomfortable, and dealing with this is so
hurtful it's *literally* making me physically ill.

yet... i have a goal to achieve, the target's been set, to create a
completely libre, open and freely-available 3D-capable processor, and
to *develop* it in a truly libre and open fashion, and i am not going
to go away until that goal is achieved.

so, we have at least two more years of this. that's the reality.
so, we can use these forums for the benefit of RISC-V, or not. which
is it going to be?

we can work together, to improve RISC-V and make a conscious decision
to make the forums a welcoming place... or we can turn it into a
hate-filled forum where students, corporate employees and researchers
treat it - and RISC-V - as an absolute "No Go Zone", and psychology
researchers use the public archives in case-studies over the next
couple of decades, on how open communities go into denial when faced
with challenges.

that's what we're looking at, here. it's up to everyone to make a
conscious decision which direction they want to go.


> These people don't talk to each other in the manner in which you reply to them.

people don't talk to me on comp.arch or libre-riscv-dev in the manner
in which they reply to me on here. why?

l.
Alex Bradbury's profile photo

Alex Bradbury

unread,
Nov 11, 2018, 2:05:01 PM11/11/18

Reply to author

Sign in to reply to author

Forward

Sign in to forward

Delete

You do not have permission to delete messages in this group

Link

Report message as abuse

Sign in to report message as abuse

Show original message

Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message

to Luke Kenneth Casson Leighton, [email protected]

Hi Luke,

I've changed the email subject to split this out to a separate thread.
I totally accept that your concerns about process are intertwined with
the technical discussion, but I'd much rather keep that separate to
the thread I started (more on that below). I'm not telling you what to
do, just asking.

For anyone who wants to follow the precursor to this split thread, see
<https://groups.google.com/a/groups.riscv.org/d/msg/sw-dev/dO1gR0bKfKs/z9wrlyTJAgAJ>
and <https://groups.google.com/a/groups.riscv.org/d/msg/sw-dev/dO1gR0bKfKs/zq8zanPWAgAJ>.
I don't think that the closed working groups used by the RISC-V
Foundation are the right model. This is an opinion I've expressed
repeatedly, going back to before the formation of the Foundation.
Others obviously feel differently. I think the pragmatic way forwards
is to enthusiastically embrace opportunities for more open discussion
and development. I'll note that:

1) the privileged spec source is on github with public issue tracking
<https://github.com/riscv/riscv-isa-manual>
2) the draft SBI spec is public on GitHub with public issue tracking
<https://github.com/riscv/riscv-sbi-doc/blob/master/riscv-sbi.md>
3) the de facto SBI reference implementation is upon on GitHub with
public issue tracking <https://github.com/riscv/riscv-pk>
4) The proposal from Atish on extending the SBI and all subsequent
discussions have been on public mailing lists

Given all of the above, wouldn't it make sense to work hard to keep
SBI-related technical discussions on track, with the hope of pointing
to what is hopefully a positive and focused discussion as an example
of the benefits of developing + discussing these standards out in the
open?

I'm sorry you feel that I didn't fully listen to your concerns. Mostly
there's just not much I can say in response - I can't speak for the
the RISC-V Foundation or its collaboration/development model. I was
careful not to label your concerns as invalid - I'm simply suggesting
that it would be more appropriate to raise them in a different thread,
and that you might reflect on whether the approach you're taking to
raising them is the best strategy to achieve your desired outcome.
Just a thought, and you're welcome to disagree.

Best,

Alex
Richard W.M. Jones's profile photo

Richard W.M. Jones

unread,
Nov 11, 2018, 3:31:51 PM11/11/18

Reply to author

Sign in to reply to author

Forward

Sign in to forward

Delete

You do not have permission to delete messages in this group

Link

Report message as abuse

Sign in to report message as abuse

Show original message

Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message

to Alex Bradbury, [email protected]

On Sat, Nov 10, 2018 at 09:25:29PM +0000, Alex Bradbury wrote:
> Based on a recent software working group call, it seems there is a
> desire for the SBI to ensure that a single kernel works across all
> RISC-V implementations,

Not sure what the "software working group call" is, but if this is in
any way related to what I said in my talk in Barcelona then I'd like
to clarify that I only meant for server platforms. Embedded folk can
and most likely will go their own way.

BTW the way you described the full scope of the (potential) SBI made
it sound like UEFI.

Rich.

--
Richard Jones, Virtualization Group, Red Hat http://people.redhat.com/~rjones
Read my programming and virtualization blog: http://rwmj.wordpress.com
virt-df lists disk usage of guests without needing to install any
software inside the virtual machine. Supports Linux and Windows.
http://people.redhat.com/~rjones/virt-df/
Christoph Hellwig's profile photo

Christoph Hellwig

unread,
Nov 11, 2018, 4:40:55 PM11/11/18

Reply to author

Sign in to reply to author

Forward

Sign in to forward

Delete

You do not have permission to delete messages in this group

Link

Report message as abuse

Sign in to report message as abuse

Show original message

Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message

to Richard W.M. Jones, Alex Bradbury, [email protected]

On Sun, Nov 11, 2018 at 11:31:47AM +0000, Richard W.M. Jones wrote:
> On Sat, Nov 10, 2018 at 09:25:29PM +0000, Alex Bradbury wrote:
> > Based on a recent software working group call, it seems there is a
> > desire for the SBI to ensure that a single kernel works across all
> > RISC-V implementations,
>
> Not sure what the "software working group call" is,

I have no idea either. But more importantly having important points
only on a call is always a bad idea for a truely open project. Time
zones will get in the way of attendance, nevermind that a of people
who could contribute on mailing lists simply don't have the time to
attend regularly scheduled calls for everything they are involved in.


> but if this is in
> any way related to what I said in my talk in Barcelona then I'd like
> to clarify that I only meant for server platforms. Embedded folk can
> and most likely will go their own way.
>
> BTW the way you described the full scope of the (potential) SBI made
> it sound like UEFI.

The SBI has always been described as for Unix class systems, which
was to entail more than servers, but basically everything running
Linux, FreeBSD, etc.

Christoph Hellwig's profile photo

Christoph Hellwig

unread,
Nov 11, 2018, 4:49:32 PM11/11/18

Reply to author

Sign in to reply to author

Forward

Sign in to forward

Delete

You do not have permission to delete messages in this group

Link

Report message as abuse

Sign in to report message as abuse

Show original message

Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message

to Alex Bradbury, [email protected]

Hi Alex,

as someone who spent a fair time with the SBI interface and the
Supervsior (OS) side of it I'll try to chime in.

The SBI seems to be trying to archive two things, but in the few
bits actually implemented, and even more so in the grand unified
scheme of things.

First it seems to avoid having to architecturally specific things that
in other modern architectures is part of the privileged instruction set.

In the current "stub" SBI this includes TLB flushing, instruction
cache flushing on others than the local core, IPI handling, time CSR
handling. This part is specific to the actual cpu microarchitecture
used in a given setup.

The other bit is just general platform wire up that is handled by
a firmware interface (like UEFI) elsewhere. Example include the SBI
console or the shutdown handling in the current "stub" SBI. This part
is generally board specific.

In general the first category seems like a really bad idea in
retrospective, as we could have just made the sbi interfaces instruction
and CSRs instead, leading to an actually well defined and standalone
usable supervisor isa. The other bits could then be left to the firmware
interface of choice.

Now do we need the SBI? Without having a compiler and/or assembler that
knows about the superset of all vendor specific instructions and CSRs
we very much do unless we move more of the fast (or fast-ish) path
bits the SBI currently handles into the privilged ISA as architectural
features.


[and now I'll try to find some time to read the big flame ware. Or maybe
not and enjoy the sunday afternoon instead]

Luke Kenneth Casson Leighton's profile photo

Luke Kenneth Casson Leighton

unread,
Nov 11, 2018, 4:57:09 PM11/11/18

Reply to author

Sign in to reply to author

Forward

Sign in to forward

Delete

You do not have permission to delete messages in this group

Link

Report message as abuse

Sign in to report message as abuse

Show original message

Either email addresses are anonymous for this group or you need the view member email addresses permission to view the original message

to Alex Bradbury, [email protected]

On Sun, Nov 11, 2018 at 10:05 AM Alex Bradbury <[email protected]> wrote:
>
> Hi Luke,
>
> I've changed the email subject to split this out to a separate thread.
> I totally accept that your concerns about process are intertwined with
> the technical discussion, but I'd much rather keep that separate to
> the thread I started (more on that below). I'm not telling you what to
> do, just asking.

no it makes a lot of sense. i get caught up in these things.


> I don't think that the closed working groups used by the RISC-V
> Foundation are the right model. This is an opinion I've expressed
> repeatedly, going back to before the formation of the Foundation.
> Others obviously feel differently.

when i spoke to krste in barcelona about this, he explained that
there's people who just simply would not come forward - at all - if
the discussions were public. mostly, they're people who have worked
in U.S. Classified environments, and their knowledge and expertise is
too valuable to ignore.

i understand this... i see the logic... and then i'm faced with a
massive goal and the worry of keeping my family fed and housed when i
have absolutely no idea, month to month, whether i'll make the next
rent payment, living under the radar in a Republic of China.... and
i'm locked out of access to critical information and resources that
would *literally* save months if not years of development effort.

also, samuel's concern about the ITU-style standards development
process bleeding over into software development as well *has*
basically already happened [
https://groups.google.com/a/groups.riscv.org/d/msg/sw-dev/0fU8_ZBRcm4/49tMtyyrAwAJ
]


> I think the pragmatic way forwards
> is to enthusiastically embrace opportunities for more open discussion
> and development. I'll note that:
>
> 1) the privileged spec source is on github with public issue tracking
> <https://github.com/riscv/riscv-isa-manual>
> 2) the draft SBI spec is public on GitHub with public issue tracking
> <https://github.com/riscv/riscv-sbi-doc/blob/master/riscv-sbi.md>
> 3) the de facto SBI reference implementation is upon on GitHub with
> public issue tracking <https://github.com/riscv/riscv-pk>
> 4) The proposal from Atish on extending the SBI and all subsequent
> discussions have been on public mailing lists
>
> Given all of the above, wouldn't it make sense to work hard to keep
> SBI-related technical discussions on track, with the hope of pointing
> to what is hopefully a positive and focused discussion as an example
> of the benefits of developing + discussing these standards out in the
> open?

yeah, that sounds like a great plan. what you say reminds me of
"don't complain, do it". my friend today sent me a message, "you hear
of any vision that succeeded which was a massive of 'don'ts'???" :)


> I'm sorry you feel that I didn't fully listen to your concerns. Mostly
> there's just not much I can say in response -

it's ok, i'm sorry, i just... flipped.


> I can't speak for the
> the RISC-V Foundation or its collaboration/development model. I was
> careful not to label your concerns as invalid - I'm simply suggesting
> that it would be more appropriate to raise them in a different thread,
> and that you might reflect on whether the approach you're taking to
> raising them is the best strategy to achieve your desired outcome.
> Just a thought, and you're welcome to disagree.

appreciated. it's just... i've spent literally years of my life
doing reverse-engineering, cleaning up after the proprietary, secret,
closed-door messes of microsoft, google, and chinese criminal
copyright infringers, then having people sponge off of my effforts,
make a ton of money, and claim all the credit (that's *free software*
people claiming credit for my work). i'm absolutely sick of it,
figuratively *and* literally.

so whenever i am faced with closed doors and having to face the
prospect of reverse-engineering the intent behind a document, i'm
vividly - so vividly it *really* scares me - reminded of the times
when i've quite literally been homeless, and not being able to turn
the key in the lock of my house because i had RSI so badly as i was
down to under 65kg because i couldn't buy enough food, and too many
other instances to list, burned into my memory that i'm never, ever
going to forget.

i'm sharing this so that you don't just _think_ i'm motivated to work
towards the goal of a welcoming and thriving open RISC-V community,
you really *know* it.

so yeah, let's see what happens. those are useful links to start from, alex.

l.
Alex Bradbury's profile photo

Alex Bradbury

unread,
Nov 11, 2018, 5:16:33 PM11/11/18

Reply to author

Sign in to reply to author

Forward

Sign in to forward

[Solved]-Are C++ libs created with different versions of Visual Studio compatible with each other?-C++

If you are distributing static libraries, you may be able to distribute version-independent libraries, depending on exactly what you are doing. If you are only making calls to the OS, then you may be OK. C RTL functions, maybe. But if you use any C++ Standard Library functions, classes, or templates, then probably not.

If distributing DLLs, you will need separate libraries for each VS version. Sometimes you even need separate libraries for various service-pack levels. And as mentioned by VolkerK, users of your library will have to use compatible compiler and linker settings. And even if you do everything right, users may need to link with other libraries that are somehow incompatible with yours.

Due to these issues, instead of spending time trying to build all these libraries for your users, I'd spend the time making them as easy to build as possible, so that users can can build them on their own with minimal fuss.

Are C++ libs created with different versions of Visual Studio compatible with each other??

If you are distributing static libraries, you may be able to distribute version-independent libraries, depending on exactly what you are doing. If you are only making calls to the OS, then you may be OK. C RTL functions, maybe. But if you use any C++ Standard Library functions, classes, or templates, then probably not.

If distributing DLLs, you will need separate libraries for each VS version. Sometimes you even need separate libraries for various service-pack levels. And as mentioned by VolkerK, users of your library will have to use compatible compiler and linker settings. And even if you do everything right, users may need to link with other libraries that are somehow incompatible with yours.

Due to these issues, instead of spending time trying to build all these libraries for your users, I'd spend the time making them as easy to build as possible, so that users can can build them on their own with minimal fuss.

sbis visual c runtime error

0 Comments

Leave a Comment