Wow. Thanks!

> On Dec 9, 2015, at 18:28 , Britt Durbrow 
> <bdurb...@rattlesnakehillsoftworks.com> wrote:
> 
> Well, FWIW I happen to have an iPad Pro sitting next to me, so I just whipped 
> up the following test and was able to allocate and write to almost 3GB of RAM 
> before it started crashing on me.
> 
> #define allocationSize        (1024ULL*1024ULL*3050ULL)
> #define stride                        (4096)
> 
> - (BOOL)application:(UIApplication *)application 
> didFinishLaunchingWithOptions:(NSDictionary *)launchOptions
> {
>       
>       printf("Starting memory test...\n");
>       
>       printf("Allocating memory...\n");
>       
>       uint8_t *buffer=malloc(allocationSize);
>       
>       printf("Writing to buffer...\n");
>       
>       for(uint64_t cursor=0; cursor<allocationSize; cursor+=stride)
>               buffer[cursor]=0xAC;
> 
>       printf("Done!\n");
> 
> 
>       return YES;
> }
> 
> 
> 
>> On Dec 9, 2015, at 5:56 PM, Rick Mann <rm...@latencyzero.com> wrote:
>> 
>> 
>>> On Dec 9, 2015, at 17:37 , Jens Alfke <j...@mooseyard.com> wrote:
>>> 
>>> 
>>>> On Dec 9, 2015, at 5:17 PM, Rick Mann <rm...@latencyzero.com> wrote:
>>>> 
>>>> - By "virtual memory," I'm sure you don't mean it's swapping to disk 
>>>> (flash). Or do you? If not, how is it virtual?
>>> 
>>> All memory is virtual in any modern OS — the only thing that sees ‘real’ 
>>> memory addresses is the kernel’s VM subsystem. ‘Virtual’ just means there’s 
>>> a layer of indirection between address space and RAM.
>>> 
>>> On iOS it’s just that the address space normally allocated to apps (by 
>>> malloc, etc.) isn’t backed by a swap file. So it doesn’t get paged out, but 
>>> it’s also limited by the amount of physical RAM.
>> 
>> Ah, yes, of course. Just the address translation is happening. Yeah, I need 
>> to know if the dirty limit has increased.
>> 
>> Alternatively, if mmap-ed memory would actually live in RAM, but get around 
>> the dirty limit (assuming for the sake of argument that the user isn't 
>> running other apps), then we really could get more memory without a speed 
>> penalty via mmap().
>> 
>>> 
>>> By “how much virtual memory you can use” I believe Rick means how much 
>>> _address space_.
>> 
>> No, I need to know how much I can actually allocate and dirty.
>> 
>> 
>> -- 
>> Rick Mann
>> rm...@latencyzero.com
>> 
>> 
>> 
>> _______________________________________________
>> 
>> Cocoa-dev mailing list (Cocoa-dev@lists.apple.com)
>> 
>> Please do not post admin requests or moderator comments to the list.
>> Contact the moderators at cocoa-dev-admins(at)lists.apple.com
>> 
>> Help/Unsubscribe/Update your Subscription:
>> https://lists.apple.com/mailman/options/cocoa-dev/bdurbrow%40rattlesnakehillsoftworks.com
>> 
>> This email sent to bdurb...@rattlesnakehillsoftworks.com
> 
> _______________________________________________
> 
> Cocoa-dev mailing list (Cocoa-dev@lists.apple.com)
> 
> Please do not post admin requests or moderator comments to the list.
> Contact the moderators at cocoa-dev-admins(at)lists.apple.com
> 
> Help/Unsubscribe/Update your Subscription:
> https://lists.apple.com/mailman/options/cocoa-dev/rmann%40latencyzero.com
> 
> This email sent to rm...@latencyzero.com


-- 
Rick Mann
rm...@latencyzero.com



_______________________________________________

Cocoa-dev mailing list (Cocoa-dev@lists.apple.com)

Please do not post admin requests or moderator comments to the list.
Contact the moderators at cocoa-dev-admins(at)lists.apple.com

Help/Unsubscribe/Update your Subscription:
https://lists.apple.com/mailman/options/cocoa-dev/archive%40mail-archive.com

This email sent to arch...@mail-archive.com

Reply via email to