There's a finalizer being set on `*gmp.Int` when `SetString()` is called.
Since `newnum` is a `gmp.Int` and not a `*gmp.Int` and `newnum` is copied
into `output_array`, the pointer with the finalizer set on it is no longer
referenced and the finalizer is eligible to be run when the next GC occurs.
The document you referenced states that the name attribute is deprecated
for the form element, not for input elements.
On Fri, Mar 3, 2017 at 10:31 AM, wrote:
> Initially I intended to file this as an issue on the official Github
> repository for Go but were referred to this page. This is the fi
Did you try with the example I provided on your GH issue?
https://play.golang.org/p/u6eAphPRrk
On Wed, Feb 15, 2017 at 10:14 AM Mukund 8kmiles wrote:
> Hi ,
>
> Has anyone tried compressing > ~2.5 GB of data using golang.
>
> The following function compresses files. Files less than ~2.5 GB are
>
You say that the first works as expected, but they both print "tstst=true".
This is because: "You must use the -flag=false form to turn off a boolean
flag" (https://golang.org/pkg/flag/).
Without the "=", it's ambiguous whether "-tst false" is enabling the flag,
followed by a non-flag argument "fa
Ah, right. You need to use `-covermode=atomic` if you want it to work with
the race detector, otherwise the coverage profiling itself can cause races.
On Thu, Feb 2, 2017 at 11:49 AM, Aaron Wood wrote:
> Okay, so what's interesting is that the races only show up when I run
> go test -race -cover
Are you running your test with coverage enabled? The coverage rewriting
causes the race detector to indicate the wrong lines in the source file.
On Thu, Feb 2, 2017 at 11:33 AM, Aaron Wood wrote:
> I can't unfortunately since it's private right now :( I can show you the
> entire back off source
