Fwd: Bcache Status

Fwd: Bcache Status

[Andrew Thrift]

Hello,

I notice over the past 18 months there have been a number of patches
posted to this mailing list to resolve known issues with bcache.

I cant see that these have been merged into the main bcache git at
http://evilpiepirate.org/git/linux-bcache.git, is this actually the
case or am I mistaken?

Will these patches ever be seen in mainline ?


Thanks,




Andrew

Re: Fwd: Bcache Status

[Peter Kieser]

[-- Attachment #1: Type: text/plain, Size: 1186 bytes --]



On 2015-10-11 4:04 PM, Andrew Thrift wrote:
> I notice over the past 18 months there have been a number of patches
> posted to this mailing list to resolve known issues with bcache.
>
> I cant see that these have been merged into the main bcache git at
> http://evilpiepirate.org/git/linux-bcache.git, is this actually the
> case or am I mistaken?
>
> Will these patches ever be seen in mainline ?
Backports of critical bug fixes not being done was brought to attention 
of Greg KH in September 2014, where Jens and myself mentioned that 
bcache stable patches (data corruption, stability, etc.) were not being 
backported to the Linux stable branches. The response from the author of 
bcache was "This kind of unnecessary hostility is why I don't bother 
with stable much anymore. I get enough of this bullshit at work, thank you."

Unfortunately, the situation hasn't cahnged. bcache still will eat your 
data under certain use cases, and there is no warning and bcache still 
doesn't have a testing/experimental flag set on it. I guess the ultimate 
answer was "backport and submit a pull request yourself" but I'm not 
sure this is appropriate.

-Peter


[-- Attachment #2: S/MIME Cryptographic Signature --]
[-- Type: application/pkcs7-signature, Size: 4311 bytes --]

Re: Fwd: Bcache Status

[Vasiliy Tolstov]

2015-10-12 19:36 GMT+03:00 Peter Kieser <peter@kieser.ca>:
> Backports of critical bug fixes not being done was brought to attention of
> Greg KH in September 2014, where Jens and myself mentioned that bcache
> stable patches (data corruption, stability, etc.) were not being backported
> to the Linux stable branches. The response from the author of bcache was
> "This kind of unnecessary hostility is why I don't bother with stable much
> anymore. I get enough of this bullshit at work, thank you."
>
> Unfortunately, the situation hasn't cahnged. bcache still will eat your data
> under certain use cases, and there is no warning and bcache still doesn't
> have a testing/experimental flag set on it. I guess the ultimate answer was
> "backport and submit a pull request yourself" but I'm not sure this is
> appropriate.


Sometimes ago i'm using bcache, but i'm lucky and dont loose my data=)
But now i'm prefer to check lvm cache status and wait for stabilization.

-- 
Vasiliy Tolstov,
e-mail: v.tolstov@selfip.ru

Re: Fwd: Bcache Status

[Koen Kooi]

Op 12-10-15 om 18:36 schreef Peter Kieser:
> 
> 
> On 2015-10-11 4:04 PM, Andrew Thrift wrote:
>> I notice over the past 18 months there have been a number of patches
>> posted to this mailing list to resolve known issues with bcache.
>>
>> I cant see that these have been merged into the main bcache git at
>> http://evilpiepirate.org/git/linux-bcache.git, is this actually the
>> case or am I mistaken?
>>
>> Will these patches ever be seen in mainline ?
> Backports of critical bug fixes not being done was brought to attention of
> Greg KH in September 2014, where Jens and myself mentioned that bcache stable
> patches (data corruption, stability, etc.) were not being backported to the
> Linux stable branches. The response from the author of bcache was "This kind
> of unnecessary hostility is why I don't bother with stable much anymore. I get
> enough of this bullshit at work, thank you."
> 
> Unfortunately, the situation hasn't cahnged. bcache still will eat your data
> under certain use cases, and there is no warning and bcache still doesn't have
> a testing/experimental flag set on it. I guess the ultimate answer was
> "backport and submit a pull request yourself" but I'm not sure this is
> appropriate.

If Kent won't do it (for whatever reasons), anyone can propose the patches for
backporting to the various stable kernels.

This is a seperate problem than the "patches posted to this list don't get
into mainline at all" issue.