Re: [PATCH v2 3/4] sysctl: Disallow partial updates for erroneous sysctl vectors

On Sat, Aug 15, 2026 at 05:05:26PM +0100, Bradley Morgan wrote:
> commit_conv_vec() writes the elements one by one, so a reader can
> still see a mix of old and new values while the loop runs. Same as
> before this series, so no regression, but the changelog could say
> this doesn't make the write atomic either. Otherwise someone later
> will expect the vector to change all at once.

Seems reasonable: here is what I added to the commit message:


"
... This (The staging vector) does **not** make the write atomic as a reader
can still see a partially updated vector.

...
"

Best

Joel

> 
> The copy is limited by table->maxlen and only vectors pay for it.
> 
> Fine by me.
> 
> Reviewed-by: Bradley Morgan <[email protected]>
> Thanks!

-----BEGIN PGP SIGNATURE-----

iQGzBAABCgAdFiEErkcJVyXmMSXOyyeQupfNUreWQU8FAmqDEA8ACgkQupfNUreW
QU9zgQv/bnQ92wPIe9bSeDUP0FWKwpDTbtxn/2wHV3tzqIJArmIZzkr02Cy6qsxj
/De7twF6hou7L9+YQyf777GU45woG89gvPpGLs/uJ52Jlv/sqEObSzxXJNbeWWWs
Wozys0TZZsOAkBLwwvcBky4nw6EqSodJE+c1tFfAJFkr/8JeOk8kG0mPaWRP5KeK
+DlxLVpQTySKxVB1i/cZhNmeecG6FioVeMChDtLSKUvE96NSI/D1fngEd4I1aFTJ
8W4Ha6oeALfo2iZvSA0/cM8gMVW7M+dhY5RtIMzF8ejMNSOs2BGUXUh8rxP5cRbr
XXrDhO5SJbJsOnfj0HpAj/gQGJQMh+jmUSaCFp54251TyMaIpNH1WqxgZNmA68pA
s+qOTeRjPSVTHluyR2aArr24sKaO46j2NXGO2uS3qiNAslPE1nQHfAUX5pQQBAPV
dvsCclzMGntqE4wT8YWRI1wd37XdjtQVLHi55avd28i/lOSPjwfkmxpZBgmY9hTH
bD+zj0To
=AzpY
-----END PGP SIGNATURE-----