| Date: | Fri, 27 Mar 2009 15:11:09 +0100 |
|---|---|
| From: | Peng Deng <dengpeng@xxxxxxxxx> |
| Subject: | [Gems-users] A question to the MSI_MOSI_CMP_directory protool |
Hello, I've been trying to understand how the L1/L2 cache states change according to MSI_MOSI_CMP_directory protocol provided by Ruby. I would like to know with this protocol and in the situation that one processor (PROC1) changes a cache line shared by another processor (PROC2) and then PROC2 reads the same cache line, if the L2 cache will issue a forced downgrade to PROC1's L1 cache and provides the updated data to PROC2 (so that an access to physical memory can be avoid)? Regards, P.D. |
| [← Prev in Thread] | Current Thread | [Next in Thread→] |
|---|---|---|
| ||
| Previous by Date: | Re: [Gems-users] Opal simulation question, Philip Garcia |
|---|---|
| Next by Date: | Re: [Gems-users] Opal simulation question, Berkin Ozisikyilmaz |
| Previous by Thread: | Re: [Gems-users] (no subject), itecgo |
| Next by Thread: | [Gems-users] Backoff code in LogTM, Mohammad Momin Ansari |
| Indexes: | [Date] [Thread] |