-
-
Notifications
You must be signed in to change notification settings - Fork 32.3k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
[core] perf: sx #44254
base: master
Are you sure you want to change the base?
[core] perf: sx #44254
Conversation
Netlify deploy previewhttps://deploy-preview-44254--material-ui.netlify.app/ @material-ui/system: parsed: +1.44% , gzip: +1.57% Bundle size reportDetails of bundle changes (Toolpad) |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@mui/utils/deepmerge
is incredibly unefficient but it's used all across the codebase and I don't have enough bandwidth to try to replace it. But it should be replaced.
@@ -606,6 +606,7 @@ export default function extendTheme(themeOptions?: CssVarsThemeOptions): Theme { | |||
getCssVar, | |||
spacing: getSpacingVal(spacing), | |||
font: { ...prepareTypographyVars(mergedScales.typography), ...mergedScales.font }, | |||
internal_cache: {}, |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I would have expected to have to add .internal_cache
to the system's createTheme
function only, but I ended up needing to add it at various locations. It would be neat to pass all our themes through that function to avoid repeating internal details in various places.
Improve
sx
performance by removing allocations.Example user code:
We get roughly 35% less rendering time for the example above which makes heavy use of the
sx
prop, so I guess user-code would see something in the 0-40% less rendering time depending on how much they use that prop.We can still get a lot more improvement by eliminating more memory allocations (I think we can get that example to at least 50% less rendering time), but the remaining changes would require more substantial work. The format of the
style({ [prop]: value, theme })
sx style handlers is expensive, we could instead use something likestyle(prop, value, theme)
, though IIUC the system props also use those so there's a bit of refactoring to do.The logic to create an empty breakpoint object for each
sx
object/subobject is also expensive, I've tried to remove it by lazily initializing the breakpoints buthandleBreakpoints
is used too much in the codebase so that's another large refactor.