( Using Web-mail so please excuse me for the email format) ________________________________________ From: Que, Simon Sent: Friday, June 25, 2010 3:31 PM To: Shilimkar, Santosh; [email protected] Cc: Kanigeri, Hari; Ohad Ben-Cohen Subject: RE: [RFC] omap: hwspinlock: Added hwspinlock driver
>Santosh, >> This reservation KCONFIG isn't necessary. HWSPINLOCK IP is not really for >> very generic use >> but specific usages where you need to protect the data between independent >> softwares >>running on different masters. >>For other Linux only side software even with dual core, kernel spinlock >>library >>is sufficient. >We will consider this. >> Are you building this for all OMAP's ?? >>Move this base address to plat/omap44xx.h and use it from there >>" OMAP2_L4_IO_ADDRESS" this should not be used anymore. Rather, use >>"ioremap and readl/writel" >Good points, we'll implement these. >> Why do you have two files ? This IP is not part of OMAP1 and all the code >> can >> go in mach-omap and can be exported from there. >Are you suggesting we use a library approach instead of a driver approach? We >want to make this expandable in the future. If something in the underlying >hardware changes with OMAP5, OMAP6, etc, we only need to update the device >>initialization file. The driver can remain unchanged. No I am not suggesting the library approach. What you have done is correct except that "you should just make it as early platform driver" using HWMOD. This isn't so difficult considering you have already created a platform driver. Only the initializations part will change. HW spin-lock doesn't have any direct clock control so it's even simpler. Regards Santosh -- To unsubscribe from this list: send the line "unsubscribe linux-omap" in the body of a message to [email protected] More majordomo info at http://vger.kernel.org/majordomo-info.html
