Programmers have traditionally used locks to synchronize concurrent access to shared data. Lock-based synchronization, however, has well-known pitfalls: using locks for fine-grain synchronization and composing code that already uses locks are both difficult and prone to deadlock. Transactional memory provides an alternate concurrency control mechanism that avoids these pitfalls and significantly eases concurrent programming. Transactional memory language constructs have recently been proposed as extensions to existing languages or included in new concurrent language specifications, opening the door for new compiler optimizations that target the overheads of transactional memory. This paper presents compiler and runtime optimizations for transactional memory language constructs. We present a highperformance software transactional memory system (STM) integrated into a managed runtime environment. Our system efficiently implements nested transactions that support both composition of transactions and partial roll back. Our JIT compiler is the first to optimize the overheads of STM, and we show novel techniques for enabling JIT optimizations on STM operations. We measure the performance of our optimizations on a 16-way SMP running multi-threaded transactional workloads. Our results show that these techniques enable transactional memory's performance to compete with that of well-tuned synchronization.

Transects of the submersible Alvin across rock outcrops in the Oregon subduction zone have furnished information on the structural and stratigraphic framework of this accretionary complex. Communities of clams and tube worms, and authigenic carbonate mineral precipitates, are associated with venting sites of cool fluids located on a fault-bend anticline at a water depth of 2036 meters. The distribution of animals and carbonates suggests up-dip migration of fluids from both shallow and deep sources along permeable strata or fault zones within these clastic deposits. Methane is enriched in the water column over one vent site, and carbonate minerals and animal tissues are highly enriched in carbon-12. The animals use methane as an energy and food source in symbiosis with microorganisms. Oxidized methane is also the carbon source for the authigenic carbonates that cement the sediments of the accretionary complex. The animal communities and carbonates observed in the Oregon subduction zone occur in strata as old as 2.0 million years and provide criteria for identifying other localities where modern and ancient accreted deposits have vented methane, hydrocarbons, and other nutrient-bearing fluids.

In classical isostatic computations, the isostatic reduction is made using the isostatic response function (the response of the earth's gravity field to a point load) derived for an assumed mechanism of compensation, as, for example, in the hypotheses of Pratt, Airy. A method is given here for the computation of this function directly from observational data, eliminating the need for assuming a compensation mechanism. If the earth is linear in its response to the crustal loading of the topography, the response of the earth's gravity field to this loading can be represented as the two‐dimensional convolution of the topography with the earth's isostatic response function. Actually, the topography is more likely to be the result of changes in density at depth than the cause, but the mathematical treatment is the same in both cases. Through transformation into the frequency domain, the convolution becomes multiplication, and one is led directly to the result that the isostatic response function is equal to the inverse transform of the quotient of the transforms of the Bouguer gravity anomaly and the topography. If one assumes local compensation, one can invert the isostatic response function to find the changes in density with depth that result in the uplift of the topography.

Programmers have traditionally used locks to synchronize concurrent access to shared data. Lock-based synchronization, however, has well-known pitfalls: using locks for fine-grain synchronization and composing code that already uses locks are both difficult and prone to deadlock. Transactional memory provides an alternate concurrency control mechanism that avoids these pitfalls and significantly eases concurrent programming. Transactional memory language constructs have recently been proposed as extensions to existing languages or included in new concurrent language specifications, opening the door for new compiler optimizations that target the overheads of transactional memory. This paper presents compiler and runtime optimizations for transactional memory language constructs. We present a highperformance software transactional memory system (STM) integrated into a managed runtime environment. Our system efficiently implements nested transactions that support both composition of transactions and partial roll back. Our JIT compiler is the first to optimize the overheads of STM, and we show novel techniques for enabling JIT optimizations on STM operations. We measure the performance of our optimizations on a 16-way SMP running multi-threaded transactional workloads. Our results show that these techniques enable transactional memory's performance to compete with that of well-tuned synchronization.

By using a fast Fourier transform algorithm, the two‐dimensional transform of the Bouguer anomaly (with terrain correction) BA (kx, ky), free‐air anomaly FA ( kx, Ky), and topography H (kx, ky) on a 32‐km grid has been obtained, being based on some 80,000 actual gravity observations and on the Bouguer anomaly map of the U.S.A. The azimuthally symmetric, real function R (k) = 〈Re (BA (kx, ky)/H(kx, ky))〉, averaged over azimuth so as to minimize the geological ‘noise,’ gives the transfer function of the earth to a concentrated load. Assuming local compensation, the experimentally obtained R (k) has been inverted to give the change in density with depth (to about 400 km, a rough limit imposed by the long wavelength resolution of the data) causing the isostatic compensation. The compensation appears to be in the mantle and not related to parameters less than 30–40 km deep. Some features of this isostatic model compare favorably with recent seismic results. R (k) has also been used to obtain the gravity field from the compensation and hence to construct an isostatic anomaly map by multiplying H (kx, ky) by R (k), subtracting this from BA (kx, ky), and taking the inverse transform. The assumption inherent in these calculations, that any nonlinear relationships between the gravity field of the compensation and the topography are small, has been tested by computing the coherency between the isostatic anomaly and the (topography)n and is found to be justified. Another function R′ (k) = Re 〈(FA(kx, ky) / H(kx, ky))〉 is used to predict the gravity field due to both the topography and the compensation.